Archaeology in Syria

Harvey Weiss

American Journal of Archaeology, Vol. 101, No. 1. (Jan., 1997), pp. 97-148.

Stable URL:
http://links.jstor.org/sici?sici=0002-9114%28199701%29101%3A1%3C97%3AAIS%3E2.0.CO%3B2-C

American Journal of Archaeology is currently published by Archaeological Institute of America.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained
prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in
the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/journals/aia.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic
journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,
and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take
advantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org
Sat Mar 1 00:10:47 2008
 Archaeology in Syria
HARVEY WEISS

INTRODUCTION establishing regional syntheses and tuture research

The number of new and expanded archaeological goals. Lectures and discussions focused on issues
research projects in Syria (fig. 1) continues to grow such as population and settlement histories, soils,
at a rapid rate and reflects the new administrative subsistence, land use and technology, dynamic re-
efforts of the Directorate General of Antiquities and gional climates and their effects upon cultural de-
Museums, now under the direction of Sultan Mo- velopment, and chronological refinements and syn-
hesen. Indeed, the premier archaeological event of chronisms that have emerged over the past 20 years.
the past year was the "International Colloquium on The recent excavations at Tell Banat and Tell Beydar
the Syrian Jezireh," sponsored by the Directorate (see below) drew considerable attention for their im-
General of Antiquities and Museums, with the co. portant contributions to third-millennium studies.
operation of the Free University of Berlin and the For current reports on Syrian research, two pub-
German Archaeological Institute, held 22-25 April lications are now indispensable: the invigorated and
1996 in Deir ez-Zor. encyclopedic "Chronique des activitks archkologiques
Two aspects of this colloquium marked new direc- en Syrie," by Michel al-Maqdissi of the Directorate
tions in Syrian archaeology. First, the colloquium General of Antiquities and Museums, published in
celebrated the opening of the new National Museum Syria 70 (1993) 443-576 and 72 (1995) 159-269; and
in Deir ez-Zor.The museum, under the direction of Orient-Express, edited and published triannually by
Assad Mahmoud, houses archaeological material the Institut d'art et d'archkologie in Paris. Both pub.
from the provinces of Deir ez-Zor and Hasseke. The lications provide up-to-datebibliographies of articles
galleries of the museum feature photographs and and monographs on Syrian archaeology as well as
text panels highlighting the cultural developments listings of conferences and exhibitions.
of each period from the Palaeolithic to the Islamic, Twenty-seven Syrian archaeological reports are
as well as vitrines displaying an impressive range of presented here, covering approximately two-thirds of
meticulously labeled artifacts, most previously un- the field research undertaken in 1993-1995 (fig. I).'
available to the public and of exceptional interest The reports are listed alphabetically, as many of
to the international archaeological community. them summarize research across several distinct time
Secondly, the colloquium itself brought together periods.
for the first time some 60 Syrian, European, and
American archaeologists responsible for much of
the pioneering research carried out in the Jezireh EXCAVATIONS AND SURVEYS

over the past two decades, particularly in both Has- 'Abd. Uwe Finkbeiner, Universitat Tiibingen,
seke and Deir ez-Zorprovinces but also the regions of reports:
the middle Euphrates and the Balib. The colloquium, The third excavation campaign at Tell el-'Abd
including both thematic and chronological sessions, lasted from August through October 1994.' Since
encouraged wide-ranging discussions among archae. the water table had risen more than 4 m since the
ologists specializing in different chronological peri- preceding year, we were forced to alter our excava-
ods and representing diverse methodological orien- tion plans considerably. Some of the areas we had
tations, thereby providing a provocative forum for planned to excavate were completely under water,

1 Publication of this newsletter on archaeology in Syria, ums, Adnan Bounni, Director of Excavations, and the re-
as well as of the previous Syrian newsletters (AJA 95 [1991] gional representatives of the Directorate General for their
683-740 and AJA 98 [1994] 101-58), has been made possible generous support, advice, and collegiality. I am indebted
in part by a generous subvention from the Ministry of Cul- to the authors of these reports for their cooperation and
ture of the Syrian Arab Republic and its Minister, Her Ex- interest, and to Regan Huff for editorial assistance. Sev-
cellency Najah Attar. The editors ofAJA and I are very grate- eral reports were translated by Ulla Kasten and myself.
ful for this continuing support. For a detailed report of the first two seasons in 1992
The authors of these reports join me in thanking Sul- and 1993, see U. Finkbeiner, "Tell el-'Abd," DM 8 (1995)
tan Mohesen, Director General of Antiquities and Muse- 31-83.
97
American Journal of Archaeology 101 (1997) 97- 148
98 HARVEY WEISS [AJA 101

- Ma'arat en-Numan

S Y R I A

Fig. 1. Map of archaeological sites in Syria referred to in text

while others such as the stratigraphic step trench In addition to remains of architecture likely to
in excavation area I1 on the western slope had been be Hellenistic, the enlarged square 18/25yielded the
completely eroded by water. O n the other hand, new ground plan of a 6 x 8-m room belonging to a pa-
findings were exposed by erosion that enhanced o u r latial building. T h e western wall had fallen, but the
knowledge of the site remarkably. Plans for excava- southwestern interior corner remained intact, mak-
tion were adapted to the new topography (fig. 2). ing the reconstruction quite secure. T h e northern
Area I. Due to the considerable rise of the water and southern walls showed three building phases
table, it was impossible to excavate the older EB lev- corresponding to three floors. All three floors were
els in this area. While much of the earlier excava- clearly defined by ashes o r burnt debris and by ves-
tions were completely flooded, a few squares were sels in situ.
preserved. In the west of square 18125 the erosion Excavation in the neighboring square 19125 was
edge was so straight that we could use it as a section. extremely difficult. T h e latest remains of the EB pal-
This section revealed that the upper levels contained ace layjust underneath the surface at the north, but
a fairly large building, and on that basis we decided most of the square was disturbed by 17 large, deep
to open two excavation areas: a western one in squares pits. In the latest phase, the wall at the north end
18125-2ti and an eastern one in squares 19125-26 and of the palace with a passage to the north toward the
20125-26. city wall was still extant. Only parts of the middle
 ARCHAEOLOGY IN SYRIA

erosion edge 1994

13 14 15 16 17 18 19 20 21 22 23 24 25 26
Fig. 2. 'Abd. Topographic plan and excavation areas.

phase are congruent with the later phase. The north- walk. Both walls have deep foundations and must
ern wall runs almost parallel to the grid, and also be viewed in association with the walls of the earlier
has a passage to the north allowing access to a tower phase exposed in 1992 in square 17125. Their con-
and, possibly, to a rampart walk. In the interior room, nection with the middle phase is obvious, too, as they
four floors could be differentiated. form the foundation of the northern outer wall of
The earliest phase is represented only by two par- that phase.
allel walls in the balk between 18125 and 18126. As At least one more room must have existed further
the walls are only about 70 cm apart and the interval to the east in square 20125. It is defined by the north-
was deliberately filled with pebbles, this may be a ern wall of the later phase and by the city wall. The
double wall, possibly the substructure of a rampart two walls are bonded, thus proving that the recon-
100 HARVEY WEISS [AJA 101
struction of the palace in the later phase went hand traces of at least three phases of later use; these are
in hand with the repair and renovation of the city not yet shown on the plan but can also be dated to
wall. In contrast, the tower and bastion of the mid- the EB I11 period. All later material comes from pits
dle phase are built against the city wall from the and other disturbances.
inside. From these excavations, a picture emerges of a
The most important discovery on the outer face slowly changing residential quarter with houses often
of the city wall was caused by erosion: water had taken of a single room and small courtyards. The two lanes
away all reinforcements and laid bare the stone foun- may have been connected in phase 4; possibly the
dation of the core wall. Further to the east a sup- northern lane ended at a second gate, from which
porting wall built of large stones rested on the EB a track led around the deep wadi in the south toward
glacis. This wall is part of the Hellenistic-Roman Munbaqat, the ancient Ekalate.
fortification, which was defined in former campaigns. The city wall was exposed on top of the tell and
Well-preserved bricks in receding layers were visible at the tell's southern edge where a broad stone foun-
in the section under the later stone wall, revealing dation projected 2-3 m into the lake. The construc-
the construction method of the EB glacis.- tion of the fortification is essentially the same as in
Due to erosion, the results of this campaign can- the north: first, a mudbrick wall about 2.5 m wide
not be directly connected stratigraphically to the was erected on a high stone foundation as the core
results of earlier campaigns. All phases obviously wall. For reasons of stability, reinforcements were
belong to a single building, which, because of its situ- soon added from both the outside and inside, broad-
ation, dimensions, and construction, may confidently ening the wall to almost 10 m. In square 21115 the
be called a palace or fortress. This interpretation wall bends distinctly toward the west, reaching the
is supported by the fact that the city wall and palace bank of the Euphrates above a wadi running east
were erected at the same time, at the end of the EB to west.
I1 period or the beginning of EB 111. As the 1993 The last excavation campaign at Tell el-'Abd
season proved, the oldest floor of the city gate and yielded relatively few small finds. Terracotta figurines
the oldest floor of the palace are contemporary. are again the most numerous, followed by stone tools
Area III. In 1993, the excavation of houses A, B, and a few metal objects. A large quantity of potsherds
and C in squares 20118-19 was completed (fig. 3). was also collected, mostly of EB I11 date. According
All belonged to phase 2, including a small street that to W. Orthmann, there are many parallels to Halawa
divides A and B in the north from C in the south. B pottery, which is of a slightly earlier date. Based
House E, however, underlying C, is part of an earlier on these ceramic assemblages, a secure dating of the
phase 3, which could also be found in a few walls residential quarter from phases 2 to 4 should be
under A and B. Both phases date to the EB I11 period. possible to establish.
In 1994, neighboring areas were opened in order Ahmar. Guy Bunnens, University of Melbourne,
to examine the structure of the EB residential quar. reports:
ter stretching eastward to the city wall. House K, a The last three excavation seasons of the Melbourne
structure with several rooms, was exposed against University archaeological expedition to Tell Ahmarl
the city wall. It belongs to phase 2 along with houses Ti1 Barsib in 1993, 1994, and 1995 have been ex-
A-C. As the northern and western outer walls are tremely p r o d ~ c t i v eAs
. ~ in the past, we concentrated
continuous, access must have been from the lane to primarily on the Iron Age buildings of area C in the
the south. western part of the site (fig. 4), but investigations were
In the eastern section, the city wall leans heavily also conducted in other areas, including a strati-
to the west. House K did not withstand this pressure graphic sounding on the main tell (area S) and in
any more successfully than houses A and B. The the northern sector of the Lower City (area E). In
rooms of'all three houses had been completely filled, 1995, a team of archaeologists from Columbia Uni-
as they were in danger of collapse. Thus, all instal- versity and the Metropolitan Museum of Art in New
lations and vessels set into the floor were well pre- York, led by John Russell (Columbia), joined the
served; all movables had been taken away. In the Australian expedition and was in charge of area E
southern and eastern rooms of house K there were and the sounding in area H.

Vxcavations were made possible by a three-year grant see C. Zaccagnini, "Sulla collina rossa: Sensazionali sco.
from the Australian Research Council complemented, in perte a Tell Ahmar, I'antica Til Barsib, capitale di un po-
1993,by a grant from the Ian Potter Foundation. For a pop- tente stato aramaico nella Siria settentrionale," Archeo 10
ular account of the major finds, with color photographs, (Sept. 1995) 24-33.
 ARCHAEOLOGY IN SYRIA
20 21 n

PRIVATE HOUSES CITY WALL

care wall

0t e ~ n I o m m ( ~ ~

. m -

3-m. +.am

Fig. 3. 'Abd. Plan of area 111.

The excavation of the large Iron Age building in of occupation. The earliest phase was destroyed by
area C, called building C1, was extended to the east, a severe fire in this part of the house. In contrast
toward the modern road. A new room (XV) could to what has been observed in other parts of building
thus be explored as well as part of another room C1, remains of earlier constructions were discovered
(XVI) and the north corner of courtyard XIII. A few under room XV. Walls of a slightly different orien-
soundings were excavated down to sterile soil under tation have been exposed. Their irregular construc-
room XV. It was thus confirmed that the building tion technique may indicate that they were foun-
went through three main phases during its period dation walls. The building they were intended to
 HARVEY WEISS [AJA 101

Fig. 4. Ahmar. Plan of Iron Age buildings in area C.

support either was never built or was destroyed when gal documents written before witnesses. Three doc-
building C1 was erected. uments bear seal impressions. The texts seem to form
T~vodoorways were cut into the northeast and an archive belonging to a certain Hanni whose name
southwest walls of room XV. They had recessed door recurs in several documents. A few dates, given ac-
frames and, in the earliest phase, stone thresholds cording to the Assyrian eponym system, are pre.
as well as door sockets covered by square slabs dec- served. According to Dalley, they belong to the end
orated with concentric circles around the holes of of the Assyrian Empire, the latter part of the sev-
the doorposts. The southwest doorway was already enth century B.C. The tablets provide valuable in-
blocked during the earliest phase. formation about the date of the destruction in the
In the destruction level of the earliest phase, sev. earliest phase of building C1; it is likely that the sub-
era1 cuneiform tablets and two Aramaic tablets were sequent phases lasted until after the fall of the Neo-
found in and around the doorway between rooms Assyrian Empire.
X and XI. The texts have been prepared for publi- In room XV, the stone slab that formed the south
cation by Stephanie Dalley. Nine tablets are rather corner of the threshold of the blocked doorway in
well preserved and five consist of large fragments. the southwest wall was turned over to reveal a five-
With the exception of one economic text, all are le- line L.uwian inscription on its underside. According
19971 ARCHAEOLOGY IN SYRIA 103
to David Hawkins, who will publish the text, it is an
inscription concerning Hamiyatas, king of Masu-
wari, already known from several inscriptions. The
new stele mentions, i.a., granaries and the weather
god of Aleppo.
In the debris from the destruction in the earliest
phase, in room XV, a collection of about 10 carved
ivories was found together with a few uncarved ivory
plaques (see G. Bunnens's forthcoming article on the
carved ivories from Tell Ahmar in AJA 101:3 [1997]).
The carved ivories are extremely diverse in shape
and style. Three ivories deserve special attention.
One represents an Egyptianizing head about 6 cm
high (fig. 5). The quality of this work, strongly rem-
iniscent of the so-called Phoenician style, is excep-
tional. Another ivory is carved on both sides. It is
about 16 cm high and shows a griffin standing on a
tree on one side, and an elaborate palmette on the
other. The third ivory, about 32 cm long and 10 cm
high, depicts a procession of four figures carrying
provisions. They are led by a musician. As the ivory
is broken at both ends, the scene originally may have
been longer.
In previous seasons another structure, building C2,
had been located to the northwest of building C1. Fig. 5. Ahmar. Ivory Egyptianizing head from buildingC1.
All but the southwest end of this structure has now
been uncovered. The building is more than 35 m
long and about 25 m wide and consists of more than beginning, as an extension of C1. The absolute el-
eight rooms around the northeast, northwest, and evation of C2 is 50-100 cm higher than that of the
southwest sides of a large courtyard (VII).The largest earliest phase of C1, suggesting that C2 was built
one, room VI to the northwest, was a reception room sometime after C1. Excavation in the southwest cor-
with a line of stones down the middle, probably for ner of C1 showed, however, that building C2 must
a sliding hearth. The main entrance to room VI already have existed during the earliest phase of C1.
passed through the southeast wall. A niche with a The construction of building C2 seems to have co-
flat stone slab was built on the southwest side of the incided with a reorganization of building C1 after
doorway, inside the room. A path paved with baked its conversion into an industrial structure, possibly
bricks linked room VI with building C1 through devoted to weaving activities. The residential quar-
courtyard VII. A room of the liwan type, typical of ters were then presumably transferred to the newly
Neo-Assyrian architecture, opened into the south- erected building C2. bother indication of this re-
west corner of courtyard VII. The new excavations organization is given by the blocking, mentioned
revealed that the brick structure formerly identified above, of one of the doorways in room XV of build-
as building C3, against the southwest corner of build- ing C1.
ing C1, is in fact one of the rooms of building C2 Irrigation work in the immediate vicinity of the
(room IX). excavated part of area C, to the north, accidentally
The most impressive feature of building C2 is the revealed a large chamber tomb. Its excavation brought
mosaic pavement found in its courtyard (VII). Very to light a long underground room, measuring 4.5
well preserved, the pavement is about 14.7 x 13 m x 2.7 m, with a stepped dromos on the southeast
and consists of black and white river pebbles ar- side and a barrel vault of baked bricks. A hole had
ranged in a checkered pattern. been pierced in the northwest wall, probably by an-
The connection between buildings C1 and C2 has cient robbers. Against the northeast wall of the tomb,
been clarified. Since no rooms exist on the south- a ceramic coffin was found resting on a square mud-
east side of the central courtyard of C2, near C1, and brick structure probably intended to encase -the
two doorways give access from C1 into the courtyard coffin. The tomb was entirely empty except for a
of C2, it is clear that C2 was conceived, from the very beaked lamp on a pedestal, but in the topsoil im-
104 HARVEY WEISS [AJA 101
mediately to the north of the tomb, carved ivories Work continued in the sounding opened in area
were found. S on the eastern slope of the main tell. A huge for-
To the west of the tomb, in a pit containing rocks tification system dating to the Iron Age has been
and stones, three pieces of a basalt statue were found. recognized. A wall more than 8 m thick stood at the
The reconstructed statue, probably defaced in an- top of the tell, and a series of terraces were con-
tiquity, represents a beardless man with curling hair structed down to the base of the tell. The entire length
and clasped hands. Stylistically the statue can be of the terracing structure was about 13 m.
dated to the second half of the eighth century. The archaeological strata beneath the fortification
In area E, the existing trench was extended to an system have been partially explored. The pottery,
area roughly equal to three 10 x 10 m squares. In including Combed ware and ovoid bases with round
addition to the building already partially excavated, knobs, points to a date in the Middle or Late Bronze
now called building El, another building, called E2, Age. A burial yielded a cylinder seal of a style very
with a slightly different orientation, was discovered close to that of the Alalakh VII cylinder seals. The
to the west. Its most noticeable feature is a mosaic main scene shows a male figure, on the left, worship-
pavement of black and white pebbles in a checkered ping a nude goddess standing on an animal, while
pattern, similar to the mosaic previously excavated on the right stands an interceding goddess of the
in area E and to the large mosaic of area C, but con- so-called Lama type.
structed with longer and thinner pebbles. A column 'Atij. Michel Fortin, Universite Laval, Quebec,
base was found in place on the eastern edge of this reports:
pavement. The new buildingE2 seems to be contem- The fifth and final season of excavation by the
porary with El, dated to the end of the Neo-Assyrian Canadian expedition at Tell 'Atij took place from
period o r slightly later. April to June 1993.4We concentrated exclusively on
A test sounding in the southeast corner of area the main tell, leaving aside the necropolis situated
E was dug down to nearly 3 m below the surface with- on a secondary tell. The most important operation,
out reaching bedrock. Additional soundings in var- six 5 x 5 m squares in the middle of the main tell,
ious parts of the site attempted to determine the continued work undertaken in previous seasons.
extent and density of the occupation of Iron Age Structures uncovered in level IX at the end of the
Tell Ahmar. A series of soundings across an area 45 1992 campaign turned out to belong to two building
m long and 3-6 m wide were opened to the west of phases. They were adjacent to a courtyard that was
area C in order to find the Iron Age city wall. Al- already evident in level X below. The open space in
though the bedrock was reached in two places, no level X bordered a 3.5 x 5.8 m room with walls corre-
firm conclusion could be drawn concerning the wall. sponding to those of two rooms above. Walls from
A large patch of pis6 at the west end of the sound- the structure in level X were constructed in turn al-
ings could represent the ancient fortifications of the most exactly above those from a 4 x 5.8 m room
tell. No other architectural remains could be iden- in level XI (fig. 6). A courtyard existed at the same
tified in these soundings. A small trench in area G, location in level XI, but was of a smaller size. While
at some distance to the north of area C, also failed one of the rooms of level XI opened onto the com-
to produce cultural remains. Another sounding in pacted earth floor of the courtyard, the other one
area H, in the middle of the settlement, was placed led to an alley paved with stone slabs. Beneath these
in a long depression that may mark an ancient road structures a layer of gray fill had been used to level
going from the Lions' Gate to the Acropolis. How- the area, but only a few centimeters underneath the
ever, no stratification was identified. These observa- level XI remains, walls belonging to level XI1 emerged:
tions suggest that there may have been gardens and two grill-plan buildings measuring 5 x 3.3 m and
open areas inside the fortification walls of Ti1 Barsib. 3 x 3.2 m (fig. 7). These were related to a compacted
Only a small excavation in area F, to the south of earth courtyard in which we uncovered several
the Lions' Gate, revealed traces of several small struc- hearths and some badly preserved low walls.
tures. The discovery of a piece of roughly carved ivory The deposit forming level XI11 accumulated on
in area F may indicate that ivory was worked in this sterile soil and was compact and hard to dig. Among
part of the site. the few structures associated with this level were four

* T h i s season, like previous ones, was subsidized by the the botanical and faunal aspects o f this project appear in
Social Sciences and Humanities Research Council o f Can- Bulletin of the Canadian Society for Mesopotamian Studies 29
ada. A substantial preliminary report o n this campaign (1995).
is due to appear in Syria 62 (1995). Preliminary reports o n
19971 ARCHAEOLOGY IN SYRIA 105
low walls enclosing a small square room with a large
circular oven and a single pit dug into the natural
layers below. Another nearby pit contained ashes and
the complete skeleton of a cow. Two thick, parallel
walls, reused in the level above as a platform, might
have belonged to a large building still buried beneath
a significant accumulation of unexcavated debris.
These architectural remains, although scanty, lead
us to believe that Tell 'Atij had monumental build-
ings from the time of its foundation. Diagnostic
sherds date this foundation to the Middle or Late
Ninevite 5 period; they compare well with examples
from Leilan IIIc (2700-2800 B.C.).
Flotation of soil samples byJoy McCorriston (Uni-
versity of Minnesota) isolated carbonized grains from
various species growing in the surrounding valley.
Contrary to what we first thought, grains stored in the
silos at 'Atij were not imported from the plains to the
north. A large number of faunal remains were also
collected and are currently being analyzed by Melinda
Zeder (Smithsonian Institution). A small sector con-
sisting of two 5 x 5 m squares was opened on the
very summit of the main tell to collect soil samples
from the six upper levels partly excavated in previ-
ous seasons.A new silo containing carbonized grains
was incidentally discovered in the process. It was com-
plete and well preserved, 1.9 x 3.6 x 1.6 m, made
Fig. 6. 'Atij. Structures and courtyard of level XI. of mudbrick and plastered on the interior.

Fig. 7. 'Atij. Grill-plan buildings of level XII.

106 HARVEY WEISS [AJA 101

Fig. 8. Banat. Vie:w of Tell Banat North.

Another important discovery made in this sector turned out to be culturally sterile. A 1994 study sea-
is a 2.5-m-thick,massive mudbrick wall, which, given son and pedological investigation relating to 'Atij
its alignment, originally encircled the entire tell. This are described below under Tell Gudeda; production
wall was also exposed north of the main tell where of a final report is underway.
it was possible to trace its complete height of 4 m. Banat. Thomas McClellan and Anne Porter
Since its foundations rest on sterile soil, the wall was report:
clearly designed to defend the site from the earliest The 1994 and 1995 campaigns of the Euphrates
occupation. If, as we think, the site was a grain re- Salvage Project focused on Tell Banat and Tell Banat
distribution center, a wall was probably deemed nec- North, two components of an extended settlement
essary to defend the important buildings on the main complex on the left bank of the Euphrates in the
tell and to protect the site's economic interests. An- Tishrin Dam salvage zone.5 Occupation of the site
other result of tracing the city wall is the discovery can be divided into two phases, period 11, the mid-
that more than half of the site has been washed away third millennium, and period I, the third quarter
by the river. of the third millennium.
At the base of the southern slope of the main tell, A number of large-scale constructions have been
we opened seven 5 x 5 m squares adjacent to those exposed at the Banat site cluster. At Tell Kabir, a 13
excavated in 1986 and 1987 in which domestic struc- x 22 m building dating to the mid-third millennium
tures were found. This time we unearthed only a few or earlier is tentatively identified as a long-room
unrelated walls, forcing us to conclude that the do- temple, while monuments from both periods of oc-
mestic quarter associated with the redistribution cupation are present at Tell Banat North. An artificial
center on the main tell was not located there. Finally, gravel platform extends underneath a series of pub-
a 15 x 5 m test trench was sunk at the southern tail lic buildings and a burial complex at the village of
of the main tell in order to find out whether human Tell Banat.
occupation extended in that direction; the deposit Tell Banat North. Tell Banat North, a mound 100

5 For previous reports, see T.L. McClellan, "Banat,"in Millennium at the Tell Banat Settlement Complex: Tell Ka-
H. Weiss, "Archaeology in Syria," AJA 95 (1991) 701-702; bir,"DM 8 (1995) 125-63; and Porter, "Tell BanatTomb 1,"
A. Porter, "Kabir,"in Weiss 1991, 721-22; McClellan and DM 8 (1995) 1-50. Both campaigns were funded by the
Porter, "TheThird Millennium Settlement Complex at Tell National Endowment for the Humanities and private
Banat 1994,"Orient-Express1994:3,81-82;Porter, "TheThird donations.
 ,OGY IN SYRIA 107
m in diameter and 20 m high, is a monument con-
sisting of layers of reddish gravel and powdery white
marl (fig. 8). Within this period I structure, called
White Monument I, is an earlier version known as
White Monument 11, which has a corrugated pis&
surface. This surface has been traced almost to the
top of the modern mound. Although damaged by
erosion, the apex of the monument is small in area
and shows no sign of any special construction, thus
eliminating the possibility that a shrine surmounted
it. Some of the pottery and associated bone deposits
that established the period I date for the construc-
tion of White Monument I were found on or near
the surface of White Monument 11, providing a ter-
minus ante quem for the earlier structure.
Horizontal lines of terraces are visible in section
in the construction fill of White Monument I, but
exposure of their surfaces showed them to be rough
and irregular and they were probably covered at the
time of building. Any evidence that outer extensions
of these terraces originally formed a stepped struc-
ture, or a pis6 cover like that of White Monument
11, has been destroyed by erosion.
Tell Bamt area C. Sounding C, opened during pre-
vious excavations, was expanded in 1994 to include
all the available space between three village com-
pounds and the mosque. This area is dominated by
artificial platform C, comprised of gravel deposits Fig. 9. Banat. Tomb 7, chamber C, with doorway to cham-
at least 3.5 m deep. While this platform is still im- ber D.
perfectly understood, there is no doubt that it reflects
a major earth-moving operation, comparable in forming a narrow portico from which a stepped re-
energy expenditure to White Monuments I and 11. vetment led to the lower terrace. Building 6, which
In one place the gravel platform covers a sloping con- dates to period I, is a large edifice with thick walls
struction with a corrugated surface of pale pis6 sim- built out of sizable boulders. Although different from
ilar to that of White Monument 11. The location of building7 in plan and orientation, it reused the main
this structure, currently designated White Monu- floors of the earlier structure, and consequently its
ment 111, is too constricted by modern village build- terracing. Despite interruptions by village houses and
ings to expose further. It is stratigraphically sealed pits, traces of the walls of building 6 extend over the
by the platform, which in turn lies beneath building surface for more than 45 m north to south.
7, dated to period 11. To the southeast of buildings 6 and 7, and also
Building 7 is laid out on at least three terraces cut into the gravel platform, is tomb 7 (fig. 9), around
descending from east to west across a distance of which a series of burials ranging from the mid- to
45 m. These terraces are dug into platform C. The late third millennium B.C. were located. Tomb 7 is
architecture that once stood on the highest terrace a multichambered stone building that has all the hall-
is almost entirely eroded, with the exception of two marks of a burial place for a high-ranking person
column bases and one in situ subsurface column or family. The roof is formed by nine large limestone
foundation. O n the second terrace, a series of rooms slabs, each approximately 3 x 2 m. The stone blocks
are arranged around a 7 x 7 m floor of baked brick used in the walls were carefully cut and fitted into
tiles mortared by bitumen, from which steps lead place. Bitumen was used not only to seal the massive
on both the north and south sides to the rooms sit- roof slabs and mortar and patch the masonry of the
uated on the third terrace. Also on the second ter- walls, but also to cover the brick-tile floors. Prelim-
race, to the south, is a room with four circular de- inary analysis suggests that the tomb was constructed
pressions, ca. 50 cm in diameter, along a scalloped at the same time as building 7, that is, the mid-third
floor edge. These indicate the presence of columns millennium. The contents of the tomb are mixed,
108 HARVEY 7NEISS [AJA 101
however, including material from both Banat peri- Simple ware vessels, with large round-basedjars and
ods, indicating that the tomb was reused. deep hemispherical bowls predominating. Two Eu-
Many Plain Simple ware and Euphrates Banded phrates Banded ware vessels were found in the de-
ware vessels have been identified in a preliminary bris of kiln 3, however, and it is possible that this
analysis of the tomb's pottery. In one chamber, a niche ware was also being manufactured at Tell Banat. The
contained Syrian bottles that had once been deco- material from the excavated period I kiln seems to
rated with an elaborate design of tiny shell and stone continue the ceramic tradition of period 11, with the
beads. In the northwest corner of chamber D, a cir- addition of some new types, in particular jars with
cular alabaster tabletop leaned against the wall. From narrow necks and two tight loop handles high on
the period of reuse, a disarticulated burial was found the shoulder. Further analysis of the ceramics from
associated with a bronze pin, a gold pendant, and area D will supplement these preliminary comments.
two lapis lazuli "fly" beads as well as other objects. Evidence for domestic occupation was sought in
In chamber F of tomb 7, a wooden box positioned a sounding in the northern part of Tell Banat, area
diagonally in the center of the room had mostly dis- F. Just below the surface, a pebbled street up to 2
integrated, leaving a thin wooden core encased by m wide and running east-west was excavated in two
mud. The corners of the box were marked by bronze 10 x 10 m squares. Built adjacent to the street on
straps for each plank that once fit 2 x 2 cm staves. the north side were several rooms formed by irreg-
In the northwest corner of the chamber, we found ular thin walls of small fieldstones. Ceramics indicate
a small shallow stone bowl with deep triangular in- that this level dates to period I. Beneath these rooms,
cisions for inlay. Beside it was a small broken bottle part of a large mudbrick structure was detected.
and two lapis lazuli bottle stoppers inlaid with The presence of monumental construction, pub-
florettes of gold. The northern niche of chamber lic buildings, and a ceramic industry at Tell Banat
F was damaged by stone robbing in antiquity, but and Tell Banat North indicates that the complex be-
nevertheless contained a bronze pin with a bird head longs to a high order of settlement hierarchy and
and an ostrich egg container with a stone stopper raises questions as to the nature of social and polit-
inlaid with lapis lazuli. ical organization at the site. Tell Banat may repre-
Tell Banat areas D and E: Area D, on the western sent an autonomous state or perhaps a specialized
side of the site, was devoted to pottery production. component within a larger political system, in which
Five kilns and a series of associated buildings have case its function as a component may be cultic and
been excavated, and traces of over 15 kilns are vis- perhaps mortuary. In future seasons we will address
ible on the surface. Two types of kilns have been re- these issues.
corded: small rectangular structures that appear to Bazi. Berthold Einwag and Adelheid Otto, Deut-
date to period I, and large oval structures belonging sches Archaologisches Institut, Damascus, report:
to period 11. Two main architectural levels are asso- In the spring of 1993, excavations began at Tell
ciated with ceramic production in period 11. The Bazi in the Tishrin Dam area."he fourth season
earlier, level 2, was founded on an artificial gravel was concluded at the end of October 1995. The site
deposit and consists of a large building complex with was occupied only during the Late Bronze Age, and
regular walls and rooms modified by the addition there are no subsequent structures with the excep-
of a series of inner walls. A considerable change in tion of a Roman fortress at the top of the citadel.
organization is manifest in the subsequent levels, l a Tell Bazi is located on the Jezireh riverbank of the
and Ib. The architecture is less substantial, several Euphrates, at one of the few points where the valley
isolated rooms are found, and discrete work spaces widens to a large plain.
are created by low enclosing walls and benches. No The site consists of a fortified natural hill, the cit-
architecture has been found in association with the adel, and a lower town to the west. Various stone walls
kilns of period I. Only one period I kiln was exca- visible on the surface provided ideal conditions for
vated, because it had been dug into the collapsed the investigation of the site, which had two main
remains of a period I1 kiln. goals. First, in the citadel area we examined the for-
In period 11, the kilns of area D produced Plain tification system, buildings on the slopes of the hill

V e e B. Einwag, "Tall Bazi,"AfO 40-41 (1993-1994) 228f; a1 Sayasna (Deraa),Basam a1 Ahmad and George Dabboura
Einwaget a]., "Tall Bazi-Vorbericht iiber die Untersuchun- (Damascus), and Tom McClellan, Anne Porter, Eva Strom-
gen 1993,"DM 8 (1995) 95-124; T. McClellan and A. Porter, menger, and Neil Crimes of the neighboring Tell Banat
"Archaeological Surveys of the Tishreen Dam Flood Zone," expedition for their assistance and support.
AAS 39 (in press). We wish to thank Radwan Sharaf, Usama
19971 ARCHAEOLOGY IN SYRIA 109
and on top of the plateau, the paths, and gates. Sec- el terrace of the Euphrates, only a few meters above
ond, in the lower town, several stone walls could be the valley level. During the latest seasons the north-
traced on the surface and the floors of the houses ern third of the lower town was completely excavated.
lay just beneath the surface, providing the opportu- A north-south trench was dug 100 m through the
nity to uncover large housing areas and to reveal in- housing area.
formation about the planning of the town and the The investigations conducted so far reveal that the
function of its various sectors. This study was assisted lower town was planned and erected at one time.
by the results of the first stratigraphic soundings, After approximately 100 years, it was completely de-
which showed that the lower town consisted of one stroyed and abandoned. An unexpected threat hit
building level constructed in two phases. the inhabitants so suddenly that they apparently
The citadel. The citadel lies on a spur of the moun- saved only their lives, leaving behind all their goods.
tain range bordering the Euphrates valley, separated This sudden event enables us to study not only the
from the mountains by an artificial ditch. Numer- architectural remains but also to carry out a func-
ous stone walls covering the hillside were still visible tional and social analysis of rooms and houses and
on the surface and guided the investigation during the entire community.
the first two seasons. They turned out to be fortifica- The first excavations were started in the northeast;
tion walls surrounding the citadel on several levels, a second area was later opened in the northwest. Dur-
most of them serving mainly as retaining or terrace ing the last campaign both areas were connected,
walls facilitating occupation of the steep ground. At revealing an excavated area of more than 5,000 my.
the eastern slope of the hill the lowest wall, built of So far more than 20 houses have been completely
huge limestone blocks, was investigated over a length or partially uncovered. The settlement is oriented
of 120 m. This wall is founded on one of the natural along broad streets, 6 m wide, which were evidently
horizontal rock ridges and was constructed in sev- the first part of the town to be built. The houses are
eral sections, some of which ended in a right-angle often separated by narrow lanes that were accessible
turn up the hill. The wall's primary purpose was the only to the neighboring occupants, as suggested by
fortification of the citadel. A step trench was dug lockable doors at the entrance to the alleys. The
into the eastern slope in order to reveal information ground plan and size of the houses follow a strict
about the system of terrace walls. They are also built scheme, altered only during the second phase. The
of unfinished limestone blocks and are filled behind normal ground plan of a house consisted of a large
with small stones and stone chips. Small buildings main room with a row of three or four small square
cover the slope of the hill in a steplike fashion, using rooms arranged along the long side. Each measures
the terrace walls as the back walls of the houses. about 10 x 15 m. The lower layers of the walls are
The plateau of the citadel was originally square built of limestone, and the upper layers of mudbricks.
and served as a stronghold; the northwest corner was In the middle of the main room, an oval oven is often
later cut away for the installation of a huge open connected to a low working platform. The small
cistern. On the plateau we discovered parts of two rooms served as granaries and storage rooms and
large buildings, which, based on their plans and small were mostly filled with large storage jars, some of
finds, may be interpreted as official buildings: a bit which contained carbonized grain. Numerous metal
hilani and a temple, separated from each other by objects such as weapons (including a complete dag-
a street. During Roman times the natural defense ger, 28 cm long) and tools (sickles, knives, and awls)
of the hillside and the still existing LBA fortification were found in the houses. We found that manufac-
walls were used for the construction of a fortress to turing activity took place in the main room of the
control the valley and one of the few crossing points houses as well as in separate workshops. These main
over the Euphrates. rooms thus served both as residences and as the site
On the northern slope the access to the high pla- of small-scale industry. One household in the south-
teau was investigated. A shallow depression in the western corner specialized in the fabrication of
surface leading up toward the citadel turned out to bronze tools; a variety of molds were found there,
be the access route that was controlled by a gate half- primarily made of limestone. A small steatite mold
way up the slope. Investigations revealed that this found in the same house also indicates the manu-
was built in several phases; the gate of one phase facture ofjewelry using precious metals. In another
was strengthened by a bastion. building the production of hematite weights and
Lower town. Directly west of the citadel and ad- stone or shell jewelry is suggested by the finished
joining it lies the lower town, measuring about 200 or partly finished objects as well as by the raw ma-
x 250 m (figs. 10-11). It was built on the natural grav- terials found there.
110 HARVEY WEISS [AJA 101

Fig. 10. Bazi. Lower town, northeastern area. Two domestic quarters separated by a broad street.

Fig. 11. Bazi. Lower town, northwestern area and adjoining corners of four
houses.
19971 ARCHAEOLOGY IN SYRIA 111
A trench dug in the northeastern corner extend- In sector E on the west side of site, we exposed
ing from the terrace down to the valley level shows a Seleucid level, under which was an ED I11 official
that the lower town was not fortified. This suggests building measuring 26 x 7.5 m. This building con-
that the sudden end of this flourishing town may sists of four square rooms of identical dimensions
have been brought about by the attack of an enemy. with an entrance on the west facing toward the inte-
The aim of the north-south trench was to reveal the rior of the city. Two floors for this building have
borders of the lower town, but the southern bound- been identified. Associated with the upper floor, at
ary of the settlement has not yet been reached. The the rear of the building, was a small ED I11 private
trench proved, however, that throughout the settle- residence.
ment there existed only two phases. The next cam- In sector F on the acropolis, under a Seleucid level
paigns will concentrate on uncovering the southern and two levels of Akkadian occupation, we uncov.
parts of the lower town. ered a large ED I11 building, probably a palace (fig.
Beydar. Marc Lebeau, European Center for Up- 12). This palace was rebuilt in late ED 111, but the
per Mesopotamian Studies, reports: rebuilding followed most of the earlier plan. The
The Euro-Syrian Excavations at Tell Beydar (for- original ED I11 palace was constructed of white baked
merly the European Archaeological Mission in Syria) brick and had 17 or 18 rooms and a courtyard, all
are organized by the European Center for Upper with floors covered with thick plaster. This is a rec-
Mesopotamian Studies in coordination with the tangular building, 32 x 21 m, with an entrance in
Directorate General of Antiquities and Museums, the middle of the east facade. The building conforms
D a m a s ~ u sThe
. ~ European universities participating to the architectural traditions of contemporary south-
in the Center are the Katholieke Universiteit Leuven ern Mesopotamia. The ED IIIb reconstruction re-
(Karel Van Lerberghe), Westfalische Wilhelms- placed half of the destroyed or damaged walls of the
Universitat Munster (Joachim Bretschneider), Uni- earlier building while preserving many white brick
versidad autonoma de Madrid (Fernando Valdes), walls of the first phase. Some doors were blocked,
Universitt libre de Bruxelles (Philippe Talon), and and some walls were opened. In general the archi-
Universitt de Lille I11 (Dominique Parayre). tecture is considerably poorer than that of phase
Tell Beydar is a circular 28-ha third-millennium 1. Four tablets were found in the second ED I11 oc-
city, with a diameter of 600 m, defended by a city cupation phase.
wall with seven gates. A lower town (Beydar 11) of In sector B, we found an ED 111 domestic quarter
30 ha was built in the Mitanni period, then aban- under Hellenistic house foundations. No complete
doned and later rebuilt in Neo.Assyrian times. A few house plan is yet available, but partial plans of six
hundred meters south of the third-millennium city or seven houses have been exposed. In 1993 and 1994,
is an area occupied during the Hassuna, Halaf, and under the original floor of one house, 140 tablets
Ubaid periods (Beydar 111), identified by Antoine were retrieved (fig. 13). The associated ceramics are
Suleiman. ED IIIb Also contemporary with these are two tab-
The important "Kranzhugel" city of the third mil- lets from sector E, four from sector F, and one from
lennium has been exposed in several operations. Sec- the lower town. F. Ismail, H. Hammade, W. Salla-
tor H is located at the northern gates of the city, where berger, K. Van Lerberghe, and P. Talon are respon-
the city wall was partially destroyed by Early Dynas- sible for the decipherment and publication of the
tic I1 and I11 tombs. The interior side of the city wall Beydar tablets. The 147 Beydar texts are contempo-
had small houses and workshops of ED I date that rary with those from Ebla, Mari, Abu Salabikh, and
indicate the age of the ramparts. Sector G on the Fara.
north flank of the upper city is a step-trench that The lower town (Beydar 11) was tested in sector
exposed three levels of Seleucid occupation, under J and an associated sounding. Here walls of a Neo-
which are several levels of ED 111and ED I1 domestic Assyrian residence with baked brick foundations
structures. There is evidence for an interior rampart were exposed. On the floor of this residence were
ca. 5 m wide. numerous sherds of Nuzi ware. Although damaged

7 The Tell Beydar excavation is a Syro-European proj- Leuven (Belgium), Universit6 libre de Bruxelles, Med.
ect sponsored by the Directorate General of Antiquities Campus Program 95 (European Union), Petrofina (Bel-
and Museums, Deutsche Forschungsgemeinschaft, Vor- gium), SociPtC gCnerale de Belgique, Minist6re des affaires
derverein Tell Beydar and its sponsors, Nationaal Fonds 6trangPres (France) (1992-1993), Commission nationale des
voor Wetenschappelijk Onderzoek, Fonds national de la fouilles (France) (1992-1993), European Center for Upper
recherche scientifique (Belgium), Onderzoeksraad K.U. Mesopotamian Studies, and private donors.
112 HARVEY WEISS [AJA 101
Raised floors
I I
Lower floor

Fig. 12. Beydar. Plan of acropolis, sector F, unit 1. Early Dynastic I11 official complex, phase 2.

by the Neo-Assyrian foundations, there are traces tombs belonging to rulers ofTuttul, and also exposed
of a small Mitanni-period house, a silo, and additional the southern walls of two similar tombs that extend
residences. A contemporary child's tomb contained into the center of the mound (fig. 14). The plan of
an ivory box with animal decoration; other ivory ob- these tombs is rather uniform: a doorway in the east-
jects were also retrieved in 1995. ern wall leads into a narrow middle room, at the end
Bi'alThttul. Eva Strommenger, Museum fiir Vor-
und Friihgeschichte, Berlin, reports:
In the spring of 1995, we completed operations
begun in 1980, and we plan to publish these results
before resuming work in this area.8 During recent
excavations we concentrated our efforts on the cen-
tral mound of the city, which revealed public build-
ings with secure stratigraphy. We did not reach ster-
ile soil, and it appears that several more levels remain
to be excavated.
At the southern end of the central mound, and
also in the middle for the later periods, we have sev-
eral clearly identifiable levels, which are discussed
in chronological order. First, the remains of an ap-
parently well preserved and solidly built public build-
ing may be a palace. We only exposed one room in
its entirety. O n the south side of the room is a street,
and across this street we also found solid masonry
buildings. These structures probably belong to an
early phase of the Early Dynastic period.
In the next level, we excavated four aboveground Fig. 13. Beydar. Early Dynastic I11 cuneiform tablet.

8 The first volume, by Kay Kohlmeyer and Eva Strom- series by Manfred Krebernik and Adelheid Otto. Annual
menger, concerns the burials and is in press. Tablets and reports and preliminary studies have also been published
sealings will be published in the second volume of this in several volumes of MDOG (109, 113-19, and 121-27).
19971 ARCHAEOLOGY IN SYRIA 113
of which doors lead both north and south into large on the adjacent southwestern slope of the tell. This
rooms. The northern room is larger and has wall Pillar Building was untouched by the fire and con-
niches. Tomb 3 to the east had a high stone foun. tinued to be used through later periods. Two clay
dation and a stone pavement beneath the floor, as sealings were found, both in Akkadian style. Contem-
did the single.roomed tomb 5, which is attached to porary with this building, simple mudbrick construc-
tomb 3 at the southern end. The tombs were sep- tions were built in the open area to the east, and
arated by an alley running north-south, and were after the Pillar Building passed out of use, the build-
apparently built over time, the northern ones being ing area was also filled with these constructions,
the earliest. We therefore assume that the tombs were which consisted of several bread ovens, kilns for firing
built on demand for the not yet excavated palace ceramics and terracotta figurines, a large silo, and
still within the tell. Considering this likely connec- many richly furnished tombs.
tion to a palace and the actual plan of the tombs, In the subsequent level, the deep foundation
we see a striking similarity to the cemetery at Ur, trenches of another palace were evident. This palace
although the tombs at Ur are subterranean and the was probably planned in the Shakkanakku period,
palace has also not yet been excavated. The tombs but for unknown reasons mudbrick was laid up only
contained human and animal bones, jewelry, and to a certain height, and the palace remained unfin-
large assemblages of ceramics. Remains of furniture ished. After a brief interruption the palace was
with inlay were found in the main rooms, but all finished, based on the same ground plan. This com.
grave goods had been disturbed by robbers in an- plex reminds us of the palace at Mari and could be
tiquity who apparently entered through gaps in the described as a smaller, more modest version. An en-
outer walls. This looting had to have taken place trance hall and a large courtyard were traversed in
before the upper parts of the tombs were leveled for order to reach a rectangular audience hall with a
new construction. throne on the short eastern wall and benches for
In the next level, we clearly have a palace (fig. 15). the visitors on the longer sides. Behind this space
The excavated portion on the southern face of the was a large room with a small, cellar-like side room
central tell has a stone stairway along the slope, but to the east. A complex of buildings that could rep-
the southernmost part of the stairway has disap- resent a private residence was situated in the north-
peared due to erosion. Excavations took place around west corner of the palace. After a rather short period
a graveled courtyard surrounded by open halls o r of use this public building was abandoned and fell
rooms indicated by the remains of wooden posts. into disrepair. During the time of Shamshi-Adadl
Wood was also used for thresholds and as structural Yasmah-Adad the ruins were cleared, and a new,
support for walls. In the eastern room o r hallway, differently organized complex was built over the re-
close to the entrance to the main room, a wooden maining foundations. Doorways were added, cut
stand was found where many simple ceramic bowls through previously impassable walls. The building
were ready for use. The main room itself had at least lost its public function and was used for workshops
one woodkn pillar on a stone base that was addition- and administrative purposes. Many sealings were
ally raised by a two-tier stone threshold. At the west- found bearing the name Tuttul, the ancient name
ern outer wall we found two baths. This palace was of Tell Bi'a.
destroyed in a fire, and its contents were found largely Over this later period of the palace, many levels
in situ. These consisted mainly of simple, ordinary were found containing only simple private houses,
ceramics of which one deserves mention: the lower indicating the decline and eventual demise of Tut-
part of a two-handledjug of the depas amphikypellon tul sometime during the mid-second millennium B.C.
type, which indicates trade with Anatolia. The plan Two thousand years later a monastery was built on
and building style of the palace share definite sim- the abandoned ruins. It contained three impressive
ilarities with Palace G at Ebla, and a beautiful frag- and well-preserved mosaic floors, two of which can
ment of a relief found in the debris shows the beard be securely dated to the sixth century A.D. due to
of a composite figure that also resembles examples their AramaiclSyriac inscriptions. The length of oc-
from Ebla, all of which suggests that this level is of cupation of the monastery has yet to be determined,
comparable date. but the modern name of the site, Tell Bi'a, means
Before the final catastrophe of the Burned Palace, tell of the monastery. The only Islamic remains re-
a workshop was constructed to the west with an outer covered are the corner of a building and a number
wall fortified by pillars. Only a minor part of this of graves.
building's eastern perimeter was excavated, but the Bosra. P:M. Blanc and J.-M. Dentzer, Centre na-
far wall is visible under certain lighting conditions tional de la recherche scientifique, Paris, report:
 HARVEY WEISS [AJA 101

.- - I

I..

Fig. 14. Bi'a. View from the north of the tombs of the rulers of Tuttul. In the center
are the fully excavated tombs 3 and 4.

Fig. 15. Bi'a. The eastern hall of the Burned Palace with a stepped stone threshold at the en.
trance to the main room. In the doorway are remains of the mudbrick collapse.
 ARCHAEOLOGY IN SYRIA

Fig. 16. Bosra. Baths.

The 1995 campaign was devoted to studying ce- the west, east, and south. The steps are built of lime-
ramic material and the preparation of publications. stone slabs with a strong mortar.
Toward this end, excavation, cleaning, and comple- These walls, M.2 and M.3, resting directly upon
mentary soundings took place at the southern baths. the basin, belong to the original plan of the bath,
This permitted continued study of the zone situated and are associated with a section of floor with ma-
between the facade of the baths and the principal terial that furnishes a terminus ante quem toward
east-west street (fig. 16). the middle of the second century. Within the fill of
Area R, situated to the north of octagonal room the west part of the basin was a great quantity of
L, was reexamined in 1994 by a sounding against the rooftiles, suggesting that this area was covered. An
wall to the north. It was therefore possible to study alignment of stones uncovered to the south of the
the relationships among the different structures asso- basin could be interpreted as a supporting wall for
ciated with the baths. Wall M.2, which we had sup- a level further south, or, more likely, as the remains
posed was stopped at the north by the baths in their of a structure protecting water channels.
first stage and which had been dated to the second Wall M.6 was rebuilt in this same phase and the
half of the second century, presents a turn to the associated floor was paved with thick slabs of beige
south before attaining alignment with the western limestone. The wall, dated to the end of the fifth or
limit of room L. This wall, 1.76 m thick, was built the sixth century, served to delimit the north side
of large blocks of possibly redressed basalt. One block of the bath. A bench was set against the facade of
measured more than 1 m in height and length. The room L, cutting the limestone paving.
wall seems to have terminated at the west in a series A system of drains was found near room L. They
of three blocks, which had perhaps extended the were well constructed, perhaps at the moment when
wall and aligned it with the early facade of room this room was transformed into an octagonal heated
L. Against the southern face of this wall was a large room with a central octagonal basin. One of the
basin, 1 2 x 4.5 x 1 m, bordered by three steps on branches of the drain came out from this room to
116 HARVEY WEISS [AJA 101
the north. At 5 m from the facade it bifurcated at Previous work at the site has established its great
a right angle to the west. Apparently this drain re- importance, particularly during the Late Uruk and
ceived a conduit from the east, without doubt a bit Akkadian periods. The aim of the new project is
older. The north facade of octagonal room L was to place these important results within a broader
remodeled during the Byzantine epoch. framework by investigating material remains from
The fragmentary state of the slabs near wall M.l a wide range of periods. Research has focused on
and the facade of room L suggest that the blocks areas of Middle Uruk, Ninevite 5, Akkadian, and early
fell from a height. The pavement was removed im- second-millennium date. In all these operations we
mediately thereafter, in the Late Ommayad period, are conducting a program of contextual analysis to
in order to pave other sectors, or perhaps to be con- investigate a full range of urban contexts at Tell Brak
verted into chalk, a material particularly precious and at the Bronze Age site of Kilise Tepe in Turkey.
in a basaltic zone. To the north, within the thick mor- Roofed rooms, courtyards, streets, rubbish dumps,
tar that supported the limestone paving, we found and cesspits are being sampled in order to under-
two parallel cavities (ca. 0.5 x 0.5 m). We suppose stand better the use of space in ancient urban en-
that they are the remains of heating elements. A vironments. We have undertaken systematic wet-
period of abandonment at the end of the Late Om- sieving and flotation of remains from all types of
mayad epoch left an accumulation of cinders, per- contexts. The analysis of recovered materials, includ-
haps from baking activities, on the substructure of ing botanical and faunal remains, lithics, and pot-
the pavement and within the heating cavities. tery, is being fully integrated with micromorpholog-
As a result of this excavation, the construction of ical studies.
the baths may be securely dated to the middle of the Middle Uruk investigations. Pottery of the Middle
second century; the limestone paving documents a Uruk period, dating to the mid-fourth millennium,
new program of ambitious construction within the is found at almost all points along the mound's skirt,
baths in the fifth century. as well as in the fields beyond the main mound, sug-
A new trench was opened in 1993 between wall gesting that the settlement of Tell Brak and adjacent
M.l and the principal axis of east-west traffic in mounds was at its greatest extent during the Middle
Bosra. Work in 1993-1994 permitted the establish- Uruk, although doubts remain about the nature of
ment of a chronological sequence of occupation and early occupation in the unknown heart of the mound.
structures in this quarter. It has revealed, apart from So far we have conducted limited excavation of Mid-
traces of earlier occupation (flint and pre-Roman dle Uruk levels, exposing parts of a building in trench
ceramics), some Nabataean remains and floors sit- HS1 near the base of the large HS spur on the north-
uated at a higher level, south of the street. western edge of the mound. Finds include typical
Porticos of the Roman period were modified at Middle Uruk pottery, including fine wares, and much
a later date in the Byzantine period. A drain system flint and obsidian. Most significantly, however, we
of baked clay perhaps belongs to this period. In the recovered a collection of clay sealings bearing im-
Ayyubid-Mamlukperiod, the sector was transformed pressions of cylinder seals, some depicting proces-
by the construction of shops, a market, and a small sions of human figures. These sealings are among
drainage channel alongside the street. In 1995, we the earliest securely stratified evidence of the use
located in a parallel trench the end of wall M.2 and of cylinder seals anywhere in the Near East. Addi-
minimal traces of Nabataean floors, without, how- tional sealings with cylinder seal impressions have
ever, determining the orientation of the street dur- been found in association with Middle Uruk pottery
ing this period. This zone was modified in the Islamic on the surface of a large mound in the fields to the
period. north of the main mound.
Brak. Roger Matthews, Cambridge University, Ninevite 5 investigations. The Ninevite 5 period,
reports: covering much of the first half of the third millen-
A new program of excavations at Tell Brak in north- nium, has remained little understood in this part
eastern Syria began in 1994 and continued in 1995." of northern Mesopotamia. Areas of Ninevite 5 oc-

T u n d i n g was provided by the British School of Ar- Akkadian Hoard from Tell Brak, Syria," Cambridge Archae-
chaeology in Iraq, the McDonald Institute for Archaeolog- ologzcal Journal 4 (1994) 290-302; Matthews, W. Matthews,
ical Research, University of Cambridge, and the British and H. McDonald, "Excavations at Tell Brak, 1994," Iraq
Academy. Contextual analysis was funded by the Lever- 56 (1994) 177-94; and Matthews, "Excavations at Tell Brak,
hulme Trust. For reports of the 1994-1995 season, see R.J. 1995," Iraq 57 (1995, in press).
Matthews, "Imperial Catastrophe o r Local Incident? An
19971 ARCHAEOLOGY IN SYRIA 117
cupation have so far been investigated at three widely
separated points on the mound. Our most extensive
information has again come from the large HS spur,
where surface operations have indicated an occu-
pation spanning Early Uruk to Late Akkadian. The
earliest Ninevite 5 levels were excavated in trench
HS2 where a courtyard and adjacent rooms were ex-
posed. Cut through this level was a series of pits with
burnt fill from which we recovered Early Ninevite
5 pottery, including a rare painted chalice, and over
120 clay sealings with cylinder seal impressions,
mainly affixed to baskets and pots. Further up the
HS spur we exposed additional Ninevite 5 occupa-
tion in trench HS4. Here the earliest excavated oc-
cupation consists of a large building with multiple
plaster floors overlaid by carefully packed bricks in
two distinct blocks, one orange, the other gray. This Fig. 17. Brak. Late Ninevite 5 clay sealing from trench HS4.
building is of very early Ninevite 5 date, roughly con-
temporary with the pits and clay sealings from trench Ninevite 5 pottery was recovered, including fenes-
HS2. trated stands and vessels decorated in a strange ex-
Above this building a major change in the use of cised style. The upper building had bins, a plastered
space tzkes place, coinciding with the first appear- floor, and large quantities of bird bones in an ashy
ance of true incised Ninevite 5 pottery. Rooms are deposit on the floor, suggestinguse as a kitchen. Pot-
located adjacent to a courtyard where seven levels tery from this building included largejars, small cups,
of deposits were excavated, spanning the entirety of and a distinctive painted stand, dating to the very
the incised and excised pottery phases of Ninevite end of the Ninevite 5 period o r slightly later. Finally,
5 occupation at Tell Brak. In the courtyard deposits, in excavating area HL, a broad shoulder on the south
much of the pottery consists of cooking wares with of the mound, we identified three phases, the low-
some finer incised vessels. Of the adjacent rooms ermost consisting of a room with burnt fill from
only the two uppermost levels have so far been ex- which came an incised jar with a possible protolit-
cavated. The lower of these phases comprises a well- erate sign on one panel. Our work on these levels
constructed building of some substance, with a door- has thus established that Tell Brak covered at least
way and a plastered niche. Pottery from this phase 20 ha in the Ninevite 5 period. The existence of
includes excised vessels. From above the floor of one sophisticated administrative practices, especially the
room we recovered 350 clay sealings, bearing im- intensive use of door sealing, suggests a social and
pressions of only five different cylinder seals (fig. economic complexity previously not attributed to
17). Designs comprise ladder motifs and animals, communities of this date.
including a porcupine. In contrast to the Early Nine- Akkadian investigations. Excavation of Akkadian
vite 5 sealings from HS2, these Late Ninevite 5 seal- levels has proceeded in two areas. The first, area HP,
ings show evidence of consistent and repetitive door is a low spur at the western extremity of the mound,
sealing, suggesting the exercise of control over ac- largely made up of an enormous mudbrick structure
cess to storage rooms. The upper phase of these directly overlying a mound of late fourth-millennium
rooms yielded a good assemblage of very late Nine- date. The large structure, measuring at least 29 x
vite 5 pottery, mainly of excised style, in association 8.5 m, consists of carefully laid bricks, whose size
with Metallic ware. Above these late Ninevite 5 struc- and arrangement suggest an Akkadian date. Against
tures there is another marked change in the use of the south face of this enigmatic structure a series
space, with buildings on new alignments dating to of rooms was constructed in the Akkadian period.
the mid-third millennium B.C. From the floor of one room, with a plastered pen-
In area HF, a large spur at the northeast of the tagonal feature, came a fragmentary Akkadian in-
site, we have also investigated Ninevite 5 levels rest- scription on a clay sealing. To the west of these rooms
ing above the Uruk deposits that form the bulk of there was a massive buildup of burnt rubbish de-
the spur. Parts of two main buildings were excavated, posits from which we recovered 277 clay sealings with
the lower one containing a series of rooms with a 52 different designs of seal impressions. Scenes in-
good sequence of plastered floors and steps. Late clude banquets, chariots, contests, and a range of
118 HARVEY WEISS [AJA 101

Fig. 18. Brak. Akkadian hoard from trench HS3.

animals both mythical and real. The entire HP com- HN near the summit of the mound. Here we have
plex appears to be part of the elaborate Akkadian excavated four phases of a large building with care-
administration of the region already well attested fully plastered rooms and internal features. Painted
from previous investigations at the site. pottery indicates a date in the earlier part of the
Akkadian levels were also excavated in trench HS3 second millennium. From one level came a distinc-
near the summit of the HS spur in the course of tive bifurcated vertebra believed to belong to a
reaching the levels below. Phase 2 in this trench com- humped Zebu cow. Elsewhere, near the summit of
prised a courtyard with an adjacent suite of two the HS spur, cesspits probably of early second-
rooms with baked brick floors, an unusual feature millennium date have been excavated and intensively
at Tell Brak. At the point where a subfloor drain sampled, yielding new and important information
passed through the exterior wall of one of the rooms on diet and environment in these little understood
into an adjacent alley, a large white stone covered centuries after the demise of Akkadian rule. Next
two pots set into a hole in the floor packing. The year, work will continue on levels of immediately post-
lower of these contained a unique collection of pre- Akkadian date in area HN, as well as in all other
cious items deliberately buried as a hoard (fig. 18). parts of the site discussed above.
These were primarily silver objects, including ingots eindaris. Muhammad Khadour and Antoine
and circular or spiral lengths of rod known as an Suleiman, Directorate General of Antiquities and
early currency, but also included a well-carved Museums, Damascus and Dietrich Siirenhagen, Uni-
ImdugudlAnzu figure with a lapis body and frit and versitat Konstanz, report:
sheet gold head, and a gold plaque showing crossed Since 1993 a Syrian.German mission has carried
lions with fine incised detail. Below phase 2 we ex- out three seasons of survey work and archaeological
cavated earlier Akkadian and pre-Akkadian levels excavations in the western Afrin valley. The valley
that have been stratigraphically connected with the is situated immediately north of Jebel Sim'an some
underlying Late Ninevite 5 occupation of trench HS4, 50 km northwest of Aleppo. Geographically and cul-
thus completing an important cultural sequence turally, the valley is part of the plain of Antioch, mod-
spanning much of the third millennium. ern 'Amuq, which played an important role in rela-
Earlier second-millenniuminvestigations. A major tions between inland Syria and the eastern Aegean
aim of our work is to explore occupation at Tell Brak world and provided access to the inland trade routes
following the collapse of Akkadian rule in the re- that connected Egypt and Palestine with Anatolia
gion, and to this end we have been excavating build- as early as the fourth millennium B.C.
ing levels from the early second millennium prior The main objective of work at Tell Cindaris was
to investigating the levels below, especially in area to clarify the material culture of the region from
19971 ARCHAEOLOGY IN SYRIA 119
the second millennium B.C. through classical times. served domestic buildings and numerous disposal
The mound at Tell cindaris sits 5 km from the river, pits that were dug into the ancient slope. While the
covers an area of approximately 20 ha, and rises 20 earlier phase yielded plentiful painted pottery, in-
m above the plain. Previous archaeological investi- cluding LH IIIC imports and local imitations, the
gation in the 1930s by the Oriental Institute (Chi- later phase is characterized by the increase of red-
cago) suggested an occupation stretching from the slipped bowls and a few fragments of Late Geomet-
Halaf period to the sixth century A.D. ric imports. A span of time from ca. 1100 to 600 B.C.
There can be little doubt that the site's name dur- seems likely for these phases, although little build-
ing the Hellenistic-Roman period was Gindaros, a ing activity is documented.
place briefly described by the ancient geographer This pattern stands in marked contrast to the archi-
Strabo. For the early first millennium B.C., a tenta- tectural evidence from below the slope's surface, where
tive identification of the site with Kinalua, capital the remains of two superimposed, monumental build-
of the Neo.Hittite kingdom of Pattin, derives from ings were discovered. The stone foundation walls of
the annals of Assurnasirpal 11. During the second the upper building are 3.4 m thick, and they surround
millennium B.C., the site probably belonged to the a room with a thick lime-plastered floor. Some 40
kingdom of Mukish. m further west a glacis-like mudbrick terrace, per-
Between 1993 and 1995 three areas of the mound haps part of this complex, runs parallel to the north-
were investigated. In area I, which is located on the ern wall of the building. If the terrace is associated,
northeastern summit, 2,000 m b e r e exposed. In the entire building measured at least 65 m in length.
contrast with the low number and poor quality of Pottery finds from above the floor indicate a late
small finds, the architectural and stratigraphical evi- MB-LB date (ca. 1600-1200 B.C.). The forerunner
dence is of some importance. For the first time since of this building was traced below its northern cor-
the American excavations took place in the plain ner, where the buttressed outer facade of a mudbrick
of Antioch some 60 years ago, a sequence of build. wall resting on a massive stone foundation came to
ing levels ranging from the middle of the second mil- light, and in a small sounding dug in the southern
lennium B.C. to the Early Byzantine period can be part of the area, where a column base set on a solid
established in this region. Dating was based on ce- lime-plastered floor was exposed. The base, which
ramic finds, and also coins from the Late Roman is made of basalt and has a diameter of 1.6 m, was
and Early Byzantine periods. Stray finds of medieval probably part of a gate or doorway. Pottery finds give
Islamic pottery may indicate an occupation of the a MB I or I1 date for both the corner and the column
site, but no architectural remains from this period base (ca. 1900-1600 B.C.). A third monumental build-
have been discovered. ing some 100 m away was destroyed by later pits, but
From the Early Byzantine period a sequence of is dated by ceramics to the MB I period.
four major building levels is preserved. The latest The steep western slope of the mound and its
levels, 1 and 2, consisted of small private houses. Lev- northwestern summit are surrounded by a fortifica-
els 3 and 4 contained the stone foundation walls of tion wall with solid stone foundations. It runs ap-
a large building complex, the northern part of which proximately 325 m to the north and 100 m to the
is now eroded. It seems that it was built around a east, on a level 7 m lower than the foundations of
large courtyard of at least 450 m2.The southern lim- the MB-LB building in area I. Two steep trenches
its of this complex have not yet been traced. Finds in area I11 provided a 10-m sequence of Byzantine
of pottery, oil lamps, and coins support a date for to Iron Age levels lying above the wall. The wall it-
all Byzantine levels within the sixth century A.D. self is 8 m thick and probably Middle or Late Bronze
The fill below these buildings was brought in from Age in date, making it likely that at least during the
a distant area of the mound. It contained numerous second half of the second millennium B.C. the set-
potsherds, glass fragments, and oil lamps dating to tlement at Cindaris was a fortified town surmounted
the Late Hellenistic period (ca. 100 B.C.). This date by an acropolis in the northeastern part. By 1200
is also affirmed by a few coins and an almost com- B.C., the acropolis must have lost its former impor-
plete terracotta relief bust of Dionysos. The fill cov- tance. Only further excavation can show whether a
ers an ancient slope where a few remains of Late new center was built in the northwestern part of the
Roman walls came to light, datable by Roman Im- mound, which would be expected if an identification
perial coins (ca. A.D. 260-330). of Gindaris with Iron Age Kinalua is correct.
Two earlier phases date back to Neo-Assyrian and Gudeda. hlichel Fortin, Universitk Laval, Quebec,
Late HittitelAramaic times. Both consist of badly pre- reports:
120 HARVEY WEISS [AJA 101
The Canadian expedition conducted a fourth and
final campaign at GudedaIo in 1993, concurrently
with excavations at Tell 'Atij. Work was concentrated
in two areas. The first, located on the northern edge
of the tell, continued the 1992 excavations. In a pre-
viously uncovered square room with thick mudbrick
walls, we removed half of the plastered floor in or-
der to reach another floor some 30 cm lower. A small
test trench was sunk below this second floor level
in order to expose the wall foundations and reach
the layer underneath the walls. The material from
this test will provide us with a terminus post quem
for the room's construction. Another smaller room
nearby was further defined, and an oven and other
installations were uncovered. The most interesting
discovery in this area, however, was a 1-m-widecob-
bled alley that ran along a wall of the room just men-
tioned and continued for 7 m before turning at a
right angle northward and ending after a few meters
due to erosion. Sections made at two different points
across this alley showed that it consisted of at least
three layers of cobbles, an accumulation 80 cm deep
resulting from constant resurfacing. O n the other
side of this alley a series of plastered basins exposed
in 1987 and 1988 were removed in 1993 in order
to explore their construction. After excavating a me-
ter of debris, we came upon several superimposed
plaster surfaces. Nearby we found more modest in- Fig. 19. Gudeda. Trench along southern flank.
stallations that had been destroyed by erosion, and
their function remains elusive. ered, and due to the limited area of excavation, they
In sum, the excavation of these four aligned 5 x were difficult to interpret. Nonetheless, a square room
5 m squares on the northern edge of the tell did not with stone walls laid on sterile soil was unearthed
reveal much due to the effect of erosion in this sec- at the bottom of this trench, indicating the extent
tor. Nevertheless, the southern balks of the squares, of occupation in the south. Finally, great quantities
in places almost 2 m high, gave us a stratigraphic of sherds collected in each layer will allow us to pre-
section that supplemented the one produced by a pare a typology of ceramics at Gudeda and to trace
deep sondage opened in 1987 along the steep north- its evolution in relation to the stratigraphy. This pot-
ern flank and shaved down in 1993. We were thus tery will also be useful for dating the site, the oc-
able to draw a section showing the entire stratigra- cupation of which covered the end of Ninevite 5 and
phy of the site in the north. We hope to link the stra- perhaps the beginning of the Akkadian period.
tigraphy of this northern section to that visible in During May 1994, Michel Blackburn (Soil Depart-
the balks of six aligned 5 x 5 m squares dug along ment, UniversitC Laval),undertook several operations
the gentle southern slope of the tell. As the north- in the field to understand better the natural com-
ern and southern trenches are located along the same ponents of the tells at 'Atij and Gudeda. Ten trial
axis, we now have a master section across the entire trenches, 1 m3 each in size, were opened along a
tell from north to south. transect across the middle Habur valley where both
This long southern trench was our second main tells are located; sedimentary layers were described
area of excavation at Gudeda in 1993 (fig. 19).A suc- and sampled. Palaeosols from 'Atij and Gudeda were
cession of layers were sampled to identify vegetal also sampled, and additional samples were taken
remains, but otherwise few structures were uncov- from the levels still visible in section at 'Atij. Thin

' 0 This and previous seasons were subsidized by the So- tin, "Rapport priliminaire sur la cinquieme campagne i
cial Sciences and Humanities Research Council of Can- Tell 'Atij et la quatriime i Tell Gudeda (Printemps 1993);"
ada. A substantial preliminary report appears in M. For- Syria 72 (1995) 23-53.
19971 ARCHAEOI,OGY IN SYRIA 121
sections were made from mudbricks from 'Atij and
examined to compare their composition from one
level to another. Finally, stone fragments were also
gathered from several areas near Hasseke in an at-
tempt to pinpoint the provenance of the raw mate-
rial used for tools.
From mid-May to mid-June 1994,Fortin and a grad.
uate student conducted a study season to complete
the description of antiquities to be deposited in the
National Museum at Deir ez-Zor,thus closing the op-
erations in the field at Tell Gudeda. Various analyses
are now underway as part of the production of the
final report.
Hammam al-nrkman. Diederik Meijer, Leiden
University, reports: Fig. 20. Hammam alTurkman. MB I unbaked clay sealed
The seventh campaign of excavations at Tell Ham- bulla from administrative complex.
mam alrurkman took place between 1 May and 26
July
- . 1995. Work concentrated on the MBA adminis- years. A northern protective wall shielded a mainly
trative complex on the summit as well as on the sep- east-west building with several interconnecting
arate settlement at the extreme southern foot of the rooms. These contained, throughout the various
mound, which also dates to the Middle Bronze Age. phases, installations such as a silo and terracotta
We continued work in the same areas opened in 1992 pipes to channel rainwater to wells. The wells (two
(the sixth season), when the dual character of hab- of which have been excavated) were cylindrical holes
itation at the site was established. lined with large, superimposed pottery cylinders
During both the 1992 and 1995 seasons, excava- (diameter between 55 and 64 cm), reaching depths
tion of the administrative complex on the western exceeding 5 m.
summit of the tell (squares K-L 22-23) yielded cu- The massive mudbrick substructure of what we
neiform tablets and sealed bullae. The former in- interpret as a tower emphasizes the function of the
cluded a Mari-style economic text fragment, and the complex as a stronghold (which was also clear from
latter Old Babylonian-style as well as provincial the glacis found in 1988). A double entrance to a
Cappadocian-style sealings (fig. 20). larger room must have been guarded, since a small
The complex's earliest phase has been reached. cubicle nearby contained broken bullae, presumably
Previous work in adjacent squares had revealed that from door fastenings. Twelve of these bulla fragments
the first, MBA occupation consisted of domestic struc- were sealed; elsewhere in the complex, a complete
tures of a rather modest character; this proved also bulla was found with two sealings of the same pro-
to be true underneath the foundations of the com- vincial Cappadocian-style seal. These finds augment
plex. It is clear now that immediately after this ini- those made in 1988 and 1992, including other seals
tial, short-livedearly MB phase of domestic architec- and their ancient impressions. Together with the few
ture, the site was chosen for the important building fragments of text that were excavated, including an
that survived, with some alterations and modifica- ancient dictionary, these sealings reflect the adminis-
tions, throughout the Middle Bronze Age. It is also trative character of the building. This function was
clear that a major part of that building has disap- retained throughout its history, and is evident in sev-
peared in the erosion of the western and southern eral parts of the complex.
sides of the summit. Furthermore, it is in this area Work in the separate southern settlement was ex-
that an EB forerunner of the MB complex must have tended to four operations totaling 300 m*. The re-
been destroyed in a conflagration followed by a short moval of large quantities of overlying debris, how-
hiatus; a layer of gray and red ash with charcoal and ever, dating to our period X (the Parthian, Roman,
mixed EB and MB sherds forms a wedge between and Byzantine periods), made for slow progress. Well-
the EB and MB levels. The thin end of this wedge preserved evidence of MBA occupation, consisting
lies under the MB complex, while the thick end lies of burnt domestic architecture, was protected by a
to the west-northwest, suggesting that the structures 3-m-widemudbrick wall with buttresses. The wall con-
that produced the destruction debris were located tained a curious opening extending down vertically
in that area. and then bending north, whose function so far is
The MB complex retained its form for some 300 entirely unclear. After this wall had fallen into dis-
122 HARVEY WEISS [AJA 101

use, several young individuals were buried in the area The pre-fortification sequence of domestic struc-
before reoccupation took place. The pottery asso- tures culminated in a destruction horizon and silt-
ciated with these inhumations yields rather depend- ing prior to the construction of a fort on a different
able synchronisms with some burials on the summit alignment. Finds include a jar fragment with a
dating to the beginning of the MB I1 period: iden- cylinder-seal impression depicting a long-horned
tical pots were found. quadruped and serpent between a partially preserved
Future seasons will concentrate on the form and rosette and quadruple spiral set on a register com-
function of the dual occupation at Hammam al- posed of triangles. This impression suggests an im-
Turkman by extending and deepening especially the portant glyptic figurative tradition in the Aleppo-
southern operation. Karum-like settlements or not, Carchemish area ancestral to the better-known Ebla-
these physically separate but exactly contemporane- Hama impressions. Another, later impression was
ous settlements present a most interesting oppor- made by a square stamp with a rosette very like an
tunity to test various hypotheses concerning econ- example from Tell Mardikh IIB.
omy, hierarchy, and social relations in Syria during The fort, enclosing an area of about 300 my, was
the Middle Bronze Age. probably built as part of an integrated settlement
Jerablus-Tahtani. Edgar Peltenburg, University of plan, which included a drainage system. The unified
Edinburgh, reports: nature of this radical reorganization may be inferred
The third and fourth seasons of University of Edin- from a large drain, originating in the center of the
burgh investigations at JerablusTahtani, a small site site, that was incorporated from the beginning into
in the neighborhood of Carchemish, were carried the stone base of the fort wall. Several features of
out in 1994 and 1995." Evidence for five periods- the wall deserve mention in the context of Syrian
Late Chalcolithic (Uruk),Early Bronze Age, Late Iron EB fortification systems: it seems to have been built
Age, Hellenistic-Roman, and Islamic- has now been in curving segments; it incorporates horizontal and
recovered from a number of soundings and four upright bricks; the faces of some segments are coated
operations. Research was concentrated on the EB with a thick, white plaster; there is evidence for case-
occupations that exist 16 m above plain level in area mate construction in area I and a bastion in area
IV. Like other parts of Syria, the Tishrin area became IV; and it was dramatically altered by the addition
densely settled with defended urban centers and their of a secondary external bank some 12 m wide. This
prolific cemeteries in the mid- and late third millen- massive, white-plastered glacis changed the appear-
nium B.C. Research has focused on the elucidation ance of the site from one surrounded by a freestand-
of secondary state formation in the Middle Euphrates ing wall to one more like a large urban center with
area, which, because of more restricted arable lands ramparts. Both the construction of the imposing wall
as well as a strategic location on trade routes, devel- and the redesign of the glacis are large-scale oper-
oped differently from better-known trajectories in ations probably introduced by an external polity,
the Habur plains and at Ebla. most likely Carchemish, a city that would not have
Prior to the Early Bronze Age, in period 1,Jerablus- permitted an independent, heavily fortified site in
Tahtani was the site of a community using exclusively its immediate neighborhood. Together with Tell el-
Uruk pottery. Sedimentological analysis indicates 'Abd and Hammam alTurkman, Jerablus 2 suggests
that riverside areas were severely disturbed by fluvial that EB I11 northern Syria had small fortlike settle-
action, and that initial occupation occurred on the ments scattered between the better-known heavily
active floodplain. Evidence from areas I, 111, and IV defended cities. Stratified pit and cist graves from the
suggests that fundamental changes took place dur- fort or a slightly later period contain pottery, figu-
ing the subsequent Early Bronze Age. Most deposits rines, decorated bone, pins, daggers, silver earrings,
can be attributed to pre-fortification or fortified and beads and pendants made of gold, carnelian,
phases, but we have only reached the upper levels rock crystal, shell, and a vitreous material.
of the carlier phase, and it is not yet possible to state On the southern spur of the site (area 11),beyond
whether occupation continued uninterrupted from the presumed line of fortifications, lies monumental
the Late Uruk period. tomb 302 with satellite burials in pithoi (fig. 21). The

1 1 For details of the 1992-1994 seasons, see E. Pelten- the British Academy, the British Ambassador in Damas-
burg, "JerablusTahtani," in H. Weiss, "Archaeology in Syria," cus, the British Institute at Amman for Archaeology and
AJA 98 (1994) 106-107; and Peltenburg et al., "Jerablus- History, the British Museum, and the National Museums
'Tahtani,Syria, 1992-4: Preliminary Report,"Levant 27 (1995) of Scotland.
1-28. Principal supporters of the investigations include
19971 ARCHAEOLOGY IN SYRIA 123

Fig. 21. Jerablus. Internal view of blocked entrance to tomb 302.

tomb is similar to and roughly contemporary with Flooding occurred at least 2.5 m above the contem-
the Ti1 Barsib hypogeum. Its principal chamber is poraneous landscape. This series of inundations un-
a 6.6 x 3.5 m corbel-walled room, oriented east- doubtedly disrupted the flow of subsistence supplies
west, with a partially blocked main entrance on the from Carchemish and could conceivably have led
short, west side. It was constructed with massive, trans- to the abandonment of JerablusTahtani. It may also
versely laid, undressed limestone blocks in six support arguments for widespread climatic change
courses, with larger blocks for the entrance lintels. and general systems collapse in the Near East and
Encasing the stone components was a two-phase elsewhere in the later third millennium. These is-
mound, at least 8 x 10 x 2 m in size, which would sues require further analysis of collected samples
have been a visible monument before tell formation and continued onsite investigations.
engulfed it. The interior of the tomb contained 13 Knedii. Ralf B. Wartke, Berliner Vorderasiatische
child and adult inhumations but was thoroughly ran- Museum, reports:
sacked for precious items. What the looters left be- In 1993 the Berliner Vorderasiatische Museum be-
hind provides a glimpse of the riches that must have gan excavations at Tell Knedig supported by the
accompanied this elite group: gold and rock crystal Deutsche Forschungsgemeinschaft.The site is located
beads, silver pins and gold foil, ivory pommels as south of Hasseke about 300 m from the western bank
at Tell Bi'a, copper alloy shafthole axes, poker-butted of the Habur. This area falls within the salvage zone
spears and daggers, and many pots lying shattered soon to be flooded by the dam under construction
in discrete concentrations, including tall, heavy gob- at the time of writing but which, according to the
lets o r champagne vessels. The last items suggest Syrian Department of Antiquities, will have been
funerary rites reminiscent of the drinking feasts de- completed by the end of 1996. Ancient settlement
picted on the standard from the Royal Cemetery at covered both the tell, which at its highest point rises
Ur. The tomb complex was set off by walls. The west- 15.5 m above the plain, as well as a lower town. Field-
ern wall, in front of the entrance, borders a passage work has exposed an area of about 40,000 m2.
at least 20 m long that leads to the upper mound Three seasons of excavations have clarified the
by a sloping, stone-paved floor and seven stairs. most important questions concerning the chronol-
Geomorphological and sedimentological studies ogy of settlement as well as the temporal relation-
indicate that the Euphrates River had abnormally ship between the upper and lower towns. Our oper-
high levels in the later third millennium. The evi- ations focused on the plateau of the tell and the
dence is particularly clear around tomb 302, where southwest and northeast portions of the lower town.
fluvial gravels were deposited in a series of deep chan- A stratigraphic section on the eastern slope of the
nels cut into the repeatedly refurbished mound. tell allowed the determination of the settlement
124 HARVEY WEISS [AJA 101

Fig. 22. Knedig. View of the lower town, with two EB buildings in foreground.

sequence. Strata exposed from the plateau down buildings. Single-and double-pot burials were located
to the lower town could be grouped into 15 con- within the walls, so that findings from the rooms were
struction levels: level 1 dates to the Islamic period; closely associated with burials. A very similar pat-
level 2 has only sparse remains but on ceramic evi- tern exists at Tell Boueid, a Syrian excavation under
dence, chiefly Roman Bitumen ware, is dated to the direction of Antoine Suleiman, where burials
theRoman-PartXian period;levels 3-6 belong to the were placed within rooms and parallel to the Neo-
Neo.Assyrian period of the eighth and seventh cen- Assyrian house walls. These burials at Knedig within
turies B.C.; and building levels 7-15, which comprise an occupied house are distinguished from the graves
more than 10 m of stratigraphy, are dated to the Early of level 6a, which were in some cases intrusive into
Bronze Age and represent the most important set- the mudbrick walls of level 6b, and also from an older
tlement period of the tell. concentration of graves in level 6b, which were bur-
O n the plateau of the tell, the Islamic construc- ied in mudbrick rubble and roof collapse. In these
tion consists of small residences, which stop abruptly cases the buildings had already been abandoned at
at the wall along the eastern edge of the tell. The the time of the burials.
remains of the Roman-Parthian level 2 wall lie di- As expected, Bronze Age remains (level 7) were
rectly on top of the latest Neo-Assyrian remains. found very close to the surface in the northern part
Among these Iron Age levels, only level 6 was broadly of the plateau. A severe Iron Age disturbance is still
exposed, within which two phases could be distin- without explanation. Excavation revealed a small,
guished. These construction phases are characterized 2.25 x 2.25 m room in which a tannur, traces of
by large structures with rectangular rooms on the plaster in a corner, and additional graves in mud-
order of 3.7 x 11.2 m in size. NumerousNeo-Assyrian brick cists were found. An additional rectangular
burials were contemporary with these mudbrick room was uncovered that was filled with mudbrick
19971 ARCHAEOL.OGY IN SYRIA 125

debris, numerous ceramic fragments, and the skel-

etal remains of an individual who died when the
building collapsed. The settlement remains from level
7 are dated by ceramics to the Early Dynastic III-
Early Akkadian period; the assemblage included
diagnostic types such as long-necked flasks and Me-
tallic ware vessels with twin handles.
In the southwestern part of the lower town, exten.
sive El3 architectural remains were located at the foot
of the tell slope (fig. 22). These had been disturbed
in recent decades, however, by round silos dug by
the seminomadic inhabitants of the surrounding
area. One interesting feature is a 32-m-long mud-
brick wall running northeast to southwest, dated to
the Neo-Assyrian period but with a foundation set
in EB levels. At both ends the wall makes a sharp
turn to the northeast, but its function and extent Fig. 23. Knedig. Early EB painted ceramic funerary vessel,
remain unclear. Excavation in the northeastern part similar to examples from Leilan IIIc.
of the lower town proved more productive. A notice-
able rise at the edge of the settled area seemed a often covered with mudbricks and large ceramic frag-
promising spot to investigate. Three Neo-Assyrian ments. One interesting Neo-Assyrian burial made
construction levels (1'-3') could be distinguished here, use of an EB pithos. In some cases the double pots
but all were severely disturbed by silos as well as by contained multiple burials, and evidence also exists
recent Islamic burials. The EB level 4', however, was for secondary burials. Some interments were rich
broadly exposed. Two structures, a well, and a sur- in grave goods. In addition to the burial vessel itself,
rounding wall were uncovered in addition to several the dead were buried with bronze or ceramic dishes,
graves. Both buildings were oriented precisely east- iron knives and arrowheads, and items of personal
west. Building I, which measured 12.6 x 6.5 m, con- adornment such as bronze rings, earrings, arm-rings,
sisted of eight rooms connected by narrow hallways. and ankle bracelets; necklaces made of semiprecious
The rooms were devoid of artifacts. stone, flint, or glass beads; and simple gold earrings
Directly to the north, building I1 was divided into and nose-rings.
seven rooms, which were modified in later phases. Of the nonfunerary finds, Islamic-period ceram-
Although the function of the two buildings cannot ics were found only on the plateau of the tell, and
yet be clarified, use as some kind of storehouse is consisted mainly of plain ware with the occasional
more likely than a residential function. Directly to comb-stripe decoration o r glaze. One particularly
the west of building I1 is a well more than 3 m in interesting fragment from the base of a bowl dis-
diameter, which when excavated contained numer- played an underglaze design of a bird with its head
ous sherds and animal bones. Preliminary examina- bent back and a thorn piercing its throat; this motif
tion suggests that the second, northernmost build- calls to mind the story of the nightingale who, for
ing is the later of the two; ceramics indicate a love of the rose, stuck himself with her thorn and
terminus ante quem of Leilan IIIc (fig. 23). Although so was united with her. TheNeo-Assyrian assemblage
most of the architecture from the northeastern lower can be closely compared to those from Assur and
town can be dated to the first half of the third cen- Nimrud. Large flasks and pithoi with decorated rims
tury B.C., the ceramics from the EB destruction level used as burial vessels are frequent. Vessels given as
of buildings I and I1 as well as the Early Dynastic grave goods include wide dishes and small bowls,
IIIb-Akkadian well fill are comparable to the assem- round flasks, round jars, knob-footed flasks, and
blage from level 7 on the plateau. steep-walled beakers. Notable in the assemblage are
From all excavations conducted, most of the re- several blue-glazedvessels, both small flasks and open
covered burials date to the Neo-Assyrian period, al- bowls. The spectrum of EB ceramics includes
though two from the southwestern part of the lower skullcap-shaped bowls, cooking pots with horizon-
town date to the Roman-Parthian period. Seven tal, triangular, o r crescent-shaped grip handles, Me-
graves date to the Early Bronze Age. With only one tallic ware vessels, notched or incised Ninevite 5
exception, the Neo-Assyrian graves consisted of ceramics, and also rare examples of painted sherds.
single- or double-pot burials and composite burials, A definite late EB IIIlIV date can be assigned to long-
 HARVEY WEISS [AJA 101

600 m2, have documented two periods of occupa-

tion, Leilan IIa and IIb, and a series of modern pits.
The entire Leilan IIa occupation (2400-2300 B.C.)
comprises part of an elaborate ritual quarter on the
acropolis first constructed in the Leilan IIId period
(2600-2400 B.C.) and extending well beyond the limit
of the 1993 exposure. The Leilan IIa excavation unit
features a central mudbrick platform, 15 x 10 m,
capped by a mudbrick altar with multiple burnt
Fig. 24. Knedig. Ninth-eighth century B.C. Neo-Assyrian plaster surfaces; portions of large, white-plastered
seal impression. rooms built against the platform's eastern face (fig.
25), including a sealed pit filled with human bones
necked flasks in the assemblage. Unique so far is the (studied by M. Zeder, Smithsonian Institution); a
find of a fragment from a beveled-rim bowl that in. ramp and a mudbrick staircase built against the
dicates settlement from the Late Chalcolithic period. platform's southern face; three rooms, one filled with
Most of the small finds were recovered from graves. carbonized grain, and another with six large storage
In Islamic contexts scattered finds included a brass vessels and two copper or bronze spearheads on its
cosmetic container, two coins, a stone knob carved floor, set against the platform's western face; aC'green
in a rosette, and fragments of glass arm.rings. Neo- space," devoid of architecture, between the south-
Assyrian levels yielded stone duck weights and two ern staircase and the western rooms; and a massive
cylinder seals. The first, made of frit, shows an archer mudbrick enclosure wall surrounding the ritual com-
attacking a winged wild sheep, while the second is plex and extending from at least the western edge
carved in black stone and depicts a man with a club of the acropolis to the northern edge.
seizing a winged bull (fig.24). Small finds from within The white-plastered walls of the Leilan IIa phase
Bronze Age settlement contexts included a mother. were truncated, and surfaces of mudbrick and thick
of-pearl pendant in the shape of a reclining cow, mud plaster were laid over them and over ashy room
terracotta figurines depicting humans and animals, deposits. Two constructions were set upon these Lei-
and flint and obsidian artifacts. Ian IIb prepared surfaces: 1) stone walls, up to 1.2
Among the previously mentioned Neo-Assyrian m wide, constructed of coarsely prepared basalt
grave goods, additional stamp and cylinder seals were boulders (some walls had three to five courses of mud-
found. The worked stamp seals, in the form of scar. brick [40 x 37 x 8 cm] upon them, and others were
abs or animals, display on the stamp surfaces Egyp. simply topped with a surface of mud plaster and large
tianizing hieroglyphs or symbols such as a scorpion, potsherds); and 2) a large mudbrick building with
eight-pointed star, or cross. Two well-carved cylinder walls three bricks thick. Portions of two rooms of
seals were also recovered: the first, made of rock this building were retrieved in the 1993 excavation
crystal, shows a bearded man between two horses, unit.
while the other, carved in black glass, depicts two Exterior surfaces of these structures covered the
bearded men facing a stylized tree of life. Another period IIa platform, leaving only the burnt altar sur-
cylinder seal shows an archer confronting a dragon. faces at floor level. In addition to the Leilan IIb ce-
The style of the seals places the graves in the ninth ramic dating of these surfaces and constructions, ad-
or eighth century, providing information for dating ditional dating is provided by a fragment of an Old
level 6. Akkadian tablet (L93-12,fig. 26), and a fragmentary
Leilan. Harvey Weiss, Yale University, reports: Old Akkadian sealing (L93-66)inscribed "Hayabum
A new extensive excavation area of some 1,500 mz the shabra-official," studied by F. Ismail, Aleppo Uni-
on the Acropolis Northwest, previously occupied by versity, and M. Van De Mieroop, Columbia University.
Leilan village houses, was opened in September- A 4-m deposit with many crosscutting storage and
October 1993.12 The new excavations, comprising garbage pits extends down from the floors of the

l 2 Research and analyses were funded in 1993 and 1995 (1993) 995-1,004; Weiss et al., "Tell Leilan, 1993,"AAS, in
by grants from Malcolm Wiener, Leon Levy, Raymond and press; M:A. Courty and Weiss, "The Scenario of Environ-
Beverly Sackler. the National Endowment for the Human- mental Degradation in the Tell Leilan Region, NE Syria,"
ities, and Yale University. Recent research reports include in N. Dalfes, G. Kukla, and Weiss eds., Third Millennium Cli-
H. Weiss et al., "The Genesis and Collapse of Third Mil. mate Change and Old World Social Collapse (Berlin 1997)
lennium North Mesopotamian Civilization,'' Science 261 107-48.
19971 ARCHAEOLOGY IN SYRIA 127

Fig. 25. Leilan. Acropolis Northwest. Period IIa room built against eastern face of
cultic platform, with pit sealed by period IIa floor; portion of period IIb brick-
topped stone wall in foreground.

Leilan village houses removed from the excavation dates, inscriptions, and sealings), administratively
area. This deposit rests upon the collapse surfaces (cylinder seals, ':sila"-bowls,and the platform and al-
of the Leilan IIblAkkadian occupation and the stra- tar in the renovated cultic quarter), and demograph-
tum marking a post-Akkadian hiatus. No occupa- ically (construction of the city wall and nucleation
tional surfaces were observed within the pit-filled of regional settlement). Subsequent excavation of this
deposit. Some pits extended a few centimeters into quarter will aim to elucidate its pre-Akkadian, in-
the Leilan IIb mudbrick constructions. digenous administrative and religious functions, its
The Acropolis Northwest excavations confirm the period IIId initial constructions, and the period IIIc
Leilan occupational sequence defined previously, conditions that preceded it and engendered third-
documenting a very large, walled, ritual quarter first millennium urbanization and state formation.
constructed during Leilan IIId (the period of sec- The third survey of the Leilan region (r = 15 km)
ondary state formation), used through Leilan IIa was undertaken in September-October 1995 with
times (the period of indigenous state consolidation), the assistance of Elena Rova, Universitg di Venezia,
and renovated in Leilan IIb times (the period of Ak- Ulla Kasten and Debra Euler, Yale University, and
kadian imperialization). The integration of Tell Lei- Richard MacNeill. The purpose of the survey was
lan within the Akkadian imperial system imposed to examine the numerous areas and sites left un-
upon the Habur plains is now documented chrono- studied by the 1987 Leilan region survey. Coverage
logically (ceramic relative chronology, radiocarbon of the Leilan region is still incomplete, but prelim-
128 HARVEY WEISS [AJA 101
M.-A. Courty (CNRS, Paris) and S. Gaffie (Institut
national agronomique, Paris-Grignon). The 60-ha
site provided evidence for growth comparable to that
seen at Leilan in the Leilan IIId-IIb periods. The
1995 survey data also confirm the subsequent Ak-
kadian centralization of political and economic con-
trol with the displacement of secondary center pop-
ulations into Leilan and other primary administrative
centers. Virtually no evidence for occupation dur-
ing the post-Akkadian period (2200-1900 B.C.), the
period of the "17 kings who lived in tents," was iden-
tified. The 19th-centuryB.C. repopulation of the re-
gion around LeilanlShubat Enlil entailed dense set-
tlement of 64 villages, providing a new perspective
on the history and economic base of Shamshi-Adad's
empire. The fluctuation of site size through time and
Fig. 26. Leilan. Acropolis Northwest. Fragment of an Old regional distributions are under analysis. The sur-
Akkadian tablet. vey will be continued in future seasons.
The Habur plains' "routes rayonnantes," observed
inary results of the 1995 survey provide the follow- from the air by W. van Liere and J. Lauffray, have
ing regional site frequencies: 49 Leilan VII-VI been much discussed and investigated several times.
(HalaflUbaid) sites, 8 Leilan V (Late UbaidlEarly T.J. Wilkinson suggested that these radiating lines
Uruk), 7 Leilan IV (Late Uruk, 3000-2900 B.C.), 5 define limited third-millennium cultivation around
Leilan IIIa (Early Ninevite 5,2900-2800), 20 Leilan sites on the Habur plains. Due to modern intensive
IIIb-c (Middle Ninevite 5, 2800-2600), 11 Leilan cultivation the lines are no longer visible on the
IIId (Terminal Ninevite 5,2600-2400), 22 Leilan IIa ground, nor on enhanced LANDSAT images. In or-
(Leilan, 2400-2300), 13 Leilan IIb (Akkadian, 2300- der, therefore, to verify them, van Liere and Lauffray's
2200), and 64 Leilan I (Habur, 1900-1700) sites. radiating lines were scanned onto LANDSAT images
These data alter significantly previous conclusions of the Leilan survey region and the sites linked by
about Leilan regional settlement. The surprisingly four systems of 19 radiating lines were then system-
dense occupation of the Leilan region during atically surveyed. In several cases the sites that are
HalaflUbaid times (approximately eight times that linked are modern villages founded in the late 1940s
of the Brak region) coincides with a dense frequency and early 1950s, with no indications of earlier settle-
of similar occupations observed in a test survey of ment. In most cases, the linked sites are neither con-
the Radd swamp around Sultan etTellul,30 km south temporaneous nor third-millennium occupations.
of Leilan, at the present-day 200-mm isohyet. These In each of the four systems' regions there are sig-
data also confirm other observations for a wetter nificant third-millennium sites that are not linked
HalaflUbaid period, dense settlement in northern by radiating lines. The radiating lines cannot, there-
Mesopotamia, and an early trend toward political fore, be third-millennium roads, and are probably
centralization at Leilan suggested by recent excava- 20th-century cart and herd tracks. Systematic inter-
tion data for a Leilan Late Ubaid settlement of ca. site surface survey also failed to produce the ceramic
20 ha. The severe reduction in Leilan regional settle- distributions typical of ancient manuring residues.
ment (both number of sites and total area occupied) Lastly, Buhairat Shura, a ca. 1-km2, perennially
during the Uruk through Early Ninevite 5 periods flooded depression west of Tell Garassa, was sam-
remains to be explained. pled by P. deMenocal, Columbia Lamont Earth Ob-
Leilan IIId, the period of regional and northern servatory, S. Leroy, Queens University, Belfast, and
Mesopotamia urbanization and secondary state for- M.-A. Courty in order to reconstruct mid-late Ho-
mation, shows the absorption of earlier Leilan IIIb-c locene regional variations in precipitation and vege-
village settlements into Leilan itself and other town tation. The depression was sampled at continuous
sites, while the succeeding period of state consoli- 2-cm intervals over a 7-m trench surface to a depth
dation shows an expansion of state control over the of 7.3 m for pollen analysis, mineralogy, grain size,
agricultural landscape. Topographic mapping, sys- magnetic properties, and radiocarbon dating in or-
tematic sherd collection, and soil testing were under- der to test for variations in sediment lithology and
taken 14.5 km west of Tell Leilan, at Tell Aid, by vegetation associated with past fluctuations in the
19971 ARCHAEOLOGY IN SYRIA 129
extent and depth of this depression. The deposit was x 4 m limestone blocks, which we called hypogeum
a succession of unconsolidated clays with variable G4. This room, located nearly 5 m below the level
quantities of coarse silt. Lithologic units were gra- of Royal Palace G, must have been a funerary space,
dational, with some boundaries indicating abrupt now completely pillaged. Built below the floorings
changes in water depth and sediment delivery. The of the Archives Palace, it was originally covered with
samples are now under analysis. stone slabs that formed a corbeled vault. This dis-
MardikhlEbla. Paolo Matthiae, Universita di covery is particularly important because it permits
Roma "La Sapienza," reports: the area of the royal tombs of the Archives period
The 30th and 31st excavation campaigns at Tell to be identified.
MardikhlEbla took place between August and No. In the Lower Town North, the enlargement of 1992
vember of 1993 and 1994.13 In 1992, in the Lower excavations in the western part of building P4 led
Town North, a 4 x 9 m sounding north of the North- to the discovery of a new short section of this multi-
ern Palace (area P) revealed an important palatial functional structure at the foot of the central hillock
building presumably of MB I date (ca. 2000-1800 of the acropolis. Building P4, the first building of
B.C.). Therefore, most of the 1993 operations were the Archives period in the Lower Town, probably
concentrated in sector P North, with the goal of ex. included workshops for the production of high-
posing this building, formerly called building P5 and quality sculpture, and was associated with an ancient
now defined as the Archaic Palace. At the same time, sacred area from the third quarter of the third mil.
the exploration of the western wing of Royal Palace lennium, on which was later built the large Sacred
G, dating to EB IVA (ca. 2400-2300 B.C.) and located Area of Ishtar, to which temple P2 and monument
on top of the acropolis, continued on a reduced scale. P3 belonged.
Area P South excavations were also resumed in EB The excavations in area P North in 1992 and 1993
IVA building P4, which was found beneath the pub- uncovered a large extension (ca. 550 mS)of the Ar-
lic square of the Sacred Area of Ishtar and between chaic Palace, the chronology of which had previously
the facades of temple P2 and the huge cult terrace been undetermined (fig. 27). An ancient floor on
called monument P3. Moreover, soundings were which sherds of Chalciform Painted ware of Hama
made at the western limits of the MB I1 Northern 55-1 and 'Amuq J types were discovered allowed us
Palace in area P West, and excavation of the balks to ascertain that the Archaic Palace was founded in
began in the northern region of the same palace. EB IVB, probably around 2150 B.C., which corre-
Finally, exploration in the later levels of area G West sponds to the Post-Palatial period of Ebla (Mardikh
continued in order to expose a larger region of the IIR2). In this first phase, the Archaic Palace had large
level contemporary with Royal Palace G in the west- rooms up to 4 x 9 m in size, with walls 1.5-2.5 m
ern area of the acropolis. thick. In the second phase, which dates to the be-
For the large early Syrian town, two major results ginning of the MB I period, floors were raised even-
were achieved, one in the western region of the acrop- ly to 0.3 or 0.4 m and smaller rooms were added with
olis, and the other in the Lower Town North. On a slightly different orientation with respect to the
Acropolis West, the problem of the large pit and original plan, particularly in the northeast. A third
east-west terracing wall identified in 1992 was solved phase of very modest use, which probably dates to
in an unexpected way. We discovered a room at the the middle or late part of the MB 1period, followed
base of the pit, built entirely from fine, squared 5 a certain amount of soil collapse probably caused

"For previous reports, see P. Matthiae, "h'ouvelles ische Archiiobgieund Altertumkunde (Wiesbaden 1994) 167-77;
fouilles a Ebla en 1987-1989," CRAI 1990,384-431; Matthiae, and Matthiae, "The Lions of the Great Goddess of Ebla:
"L'aire sacrke d'Ishtar i Ebla: Rksultats des fouilles de A Hypothesis about Some Archaic Old Syrian Cylinders,"
1990-1992," CRAI 1993, 613-62; Matthiae, "Tell Mardikh- in H. Gasche et al. eds., Cinquante-deux reflexions sur 1eProche-
Ebla (Siria),campagna di scavi 1993," Orient-Express 1994:2, Orient ancien (Louvain 1994) 329-38. s e e also the catalogue
35-38; and Matthiae, "Tell Mardikh-Ebla (Siria),campagna by Matthiae, F. Pinnock, and G. Scandone Matthiae eds..
di scavi 1994,"Orient-Express 1994:3,12-16. See also the con- Ebla. Alle origini della civilta urbana. Trenta anni di scavi in
tributions by Matthiae, "High Old Syrian Royal Statuary Siria dell'liniversita di Roma "La Sapienm" (Milan 1995). For
from Ebta," in B. Hrouda et al. eds., Von CTruknach Tuttul mention of Ebla in the economic texts of the Ur 111 Dy-
(Munich 1992) 111-28; Matthiae, "A Stela Fragment of Ha- nasty, see D.I. Owen, "Syrians in Sumerian Sources from
dad," in M.J. Mellink et al. eds., Aspects o f A r t and Iconog the Ur 111 Period," in M.W. Chavalas and J.L. Hayes eds.,
raphy:Anatolia and Its Neighbors (Ankara 1993) 389-97; Mat- New Horizons in the Study of Ancient Syria (Malibu 1992)
thiae, "Old Syrian Basalt Furniture from Ebla Palaces and 117-21.
Temples," in P. Calmeyer et al. eds., Beitrage zur altoriental-
130 HARVEY WEISS [AJA 101
19971 ARCHAEOLOGY IN SYRIA 131

by an earthquake; these collapses of the rock layer This newly discovered, important building in the
below surely led to the final abandonment of the Ar- Lower Town North of EB IVB Ebla is probably the
chaic Palace near the end of the MB I period. Later, royal palace of the period of the Ur I11 dynasty. After
the MB I1 Northern Palace was constructed over a brief, almost complete abandonment following the
the ruins of the older building. It is now clear that destruction of the city by Sargon of Akkad, settle-
the peculiar trapezoidal plan of the Northern Palace ment appears to have been limited to the northern
descends from the similar plan of the Archaic Pal- region of the town in the Archives period. The mes.
ace, the east and west perimeter walls of which de- sengers and merchants of Ebla recorded in the eco-
termined the limits of the later building. While the nomic texts of the Ur I11 dynasty, between the end
Northern Palace was in use, the area of the Archaic of Shulgi's reign and the beginning of Shu-Sin's rule,
Palace was employed at first for some installations departed from the town of the Archaic Palace of Mar-
related to metal production, and later as a waste pit. dikh IIB.
A large number of late MB I and MB I1 pottery frag- Mashnaqa. Ingolf Thuesen, University of Copen-
ments as well as other objects were discovered there. hagen, reports:
The most important is an ivory figurine of an offer- In the spring of 1994 and 1995, a Danish archaeo-
ings bearer, who carries a gazelle against his breast; logical team conducted its third and fourth seasons
this MB I1 masterpiece of ivoryworking probably of excavation of Ubaid remains at Tell Mashnaqa.
dates to 1750-1700 B.C. This work was coordinated with the substantial ex-
The original plan of the Archaic Palace of the first cavations of the high mound by a French expedition
phase can be easily reconstructed in the western and under the direction of Dominique Beyer, Universite
central regions of the building, but the modifications de Strasbourg. The 1994 expedition exposed an area
of the second and third phases conceal the original directly east of the area excavated in 1990 and 1991,
structures in the eastern region. The 33-m-longnorth- and more than doubled the total area of exposure.
ern limit of the Archaic Palace had in the western Three main strata were identified in previous sea-
region two imposing buttresses, which were prob- sons: stratum I is the first settlement at the site with
ably augmented by a third asymmetrical buttress. building remains, stratum I1 is a meter of thick
In a general evaluation, the palace seems to have greenish-gray midden deposit, and stratum I11 con-
been divided into three sectors. The westernmost tains the remains of a platform and a large rectan-
had two large halls, one 9 x 5 m and the other 10 gular kiln. All three strata also contained burials.
by at least 5.6 m in size. In the central sector, two Our functional interpretation of the stratigraphic
smaller rooms occupied the northern periphery and spatial structure of the earliest levels was mod-
while a large and important hall was built in the in- ified after the new exposures in 1994. Stratum I al-
ner region. most disappeared in the east, and strata I1 and I11
The large hall in the center of the central sector, showed no architectural features but were pitted by
which thus was exactly in the middle of the build- burials, indicating use as a cemetery. In all, 14 buri-
ing, was 9.5 m long on its east-west axis, but its even als were discovered, adding to our knowledge of the
greater length on the north-south axis could not be Ubaid burial tradition. The side-wall burial, which
determined because the hall stretches below the is a pit burial with a short wall along the body on
Northern Palace. Two important details character- one side, occurs in two variations. In one type, the
ize this room: first, a stone bench almost a meter body is placed in an open pit, which also holds
wide stretched against the whole length of the north, the stretch of wall and a few offerings such as pot-
or back, wall and featured a buttress 0.4 m deep in tery vessels. The body is then covered with soil but
the center; second, the very well preserved gypsum the top of the burial wall remains exposed, indicat-
coating on the floor stopped at a regular distance ing position and orientation of the burial (fig. 28).
of 0.3 m from the foot of the east and west walls and In the other type, the body is placed in a subterra-
1.15m from the north bench. This second peculiar- nean burial cave, and the wall is built in order to
ity leads us to believe that a coating covered the walls, cover the entrance after the body and gifts are ar-
probably wood leaning against vertical mudbricks, ranged. Burials without the side wall were also found,
remains of which were found partially in place with the body placed in an extremely flexed position.
against the west wall. We infer from the bench against Although no traces of rope were observed, the body
the north wall that this hall was the throne room appeared to have been tied up prior to the inter-
of the building, because a similar peculiarity char- ment. A few pots were sometimes found in associa-
acterized the throne room of the later Northern tion with this third type of burial.
Palace. The condition of most skeletons did not allow for
132 HARVEY WEISS [AJA 101
tum I architecture still remained unexcavated, led
to the decision to sample Ubaid levels at Mashnaqa
during a short season in the spring of 1995. Field-
work was carried out with a small staff. In this clos-
ing season, we excavated a 15 x 5 m area to the south
of the main area, where we expected to find further
remains of stratum I architecture. Part of a house
found in 1991 was excavated, and to the south the
corner of a new building appeared. This building,
unlike the others, had broad mudbrick walls, and
within it we found ashy soil rich in artifacts, includ-
ing two fragments of stone axes or picks and a clay
token. Unfortunately, much of this building has been
eroded by the river. A simple pit burial was also
found. Fieldwork focusing on the Ubaid remains at
Mashnaqa has now been completed and the process-
ing and reporting of the results are in progress.
MishrifehlQatna. Michel al-Maqdissi,Directorate
General of Antiquities, Damascus, reports:
Mishrifeh, located at the western end of the Syr-
ian steppe, at the crossroads of three interregional
routes, was one of the principal ancient centers of
western Syria. Together with the study of earlier ex-
cavations, a new multidisciplinary research program
is designed to enhance our understanding of cen-
tral Syria and its contacts with the Euphrates valley,
the Aleppo plains, and the Syrian coast. In June and
Fig. 28. Mashnaqa. Ubaid-period side-wall burial. July 1994 and in June 199.5, excavation and survey
were undertaken to prepare a new topographic plan
of the site and to determine occupational levels
safe transfer to a laboratory, so specialists in phys- around the acropolis.
ical anthropology, odontology,molecular biology, and The team first concentrated its work within the
human genetics were brought to the field in conjunc- central part of the site around the upper town (du
tion with the excavation of the burials. Studies in- Mesnil du Buisson's "colline centrale") to study the
clude attempts to determine sex, age, nutrition, and nature of occupation and establish a strategy for
diseases among the Ubaid population. In addition the coming seasons. Chantier A, situated at the west-
to observing in situ extraction of ancient DNA from ern side of the upper town northeast of the "Pillar
human bones, archaeologists from the Department of Lot," revealed a stratigraphic sequence with three
of Antiquities in Damascus as well as those working levels-Iron 11, Late Bronze, and Middle Bronze.
in the region had an opportunity to learn the theo- Chantier B, near the foot of the southeast section
retical background for the use of palaeogenetics to of the upper town, exposed coherent architectural
characterize human biology and the palaeoenviron- remains ca. 50 cm below the surface. Several walls
ment and to see the field methods demonstrated on and stones oriented north-south and east-west were
Ubaid burials at Mashnaqa. associated with a succession of tamped earth floors
Excavated material included the usual large quan- (fig. 29), the complex perhaps being industrial in
tity of potsherds, chipped flint and obsidian, and nature. Ceramics are of the Iron I1 period, with large
groundstone adzes and mortars, as well as a smaller storage jars, lamps, and small plates. A surface find
number of bone and terracotta items. Computer rec- here, however, was the head of a MB I1 figurine (fig.
ords demonstrated that, despite doubling the total 30). MB I1 tamped earth floors with varied ceramics
amount of Ubaid pottery, the relative distribution and a bronze toggle pin were found in a deep sound-
of main types and the ratio of painted to nonpainted ing in the corner of a small room with mudbrick
pottery remained constant, implying that the exca- walls.
vated corpus adequately represents the Ubaid ma- Chantier C was opened in the western part of the
terial culture at the site. This fact, along with the upper town, a few meters west of the church con-
realization that only a small area of preserved stra- structed by du Mesnil du Buisson in 1927,in an area
19971 ARCHAEOLOGY IN SYRIA 133

Fig. 29. Mishrifeh. Iron I1 buildings in chantier B.

that had been cleared of village houses in 1992.

Twenty squares in this area exposed, on the east, an
important Iron I1 occupation comprised of a multi-
roomed L-shaped house with mudbrick walls set on
stone foundations. To the west is another building,
with thick walls of mudbrick.
The two campaigns indicate the importance of
the Iron I1 occupation in the upper town and the
significant role of the city during the period of Ar-
amaean domination. Intensive collection of surface
pottery from the northern part of the upper town,
in an area disturbed by relatively recent occupation,
provided additional evidence for MB I1 and Iron I1
habitation, as well as EB IIIlIV remains.
Mozan. Giorgio Buccellati, University of Califor-
nia, Los Angeles, and Marilyn KellyBuccellati, Cali-
fornia State University, report:
The eighth season of excavations at Tell Mozan
took place in June and July 1993, and two study sea-
sons were carried out in 1994 and 1995.14 The ex-
cavations were restricted to a limited area: a sequence
of three floor deposits in the western portion of sec-
tor B of the Royal Storehouse AK, all belonging to
stratum B12 (dated to about 2300-2200 B.C.). We de-
cided on this course of action in anticipation of
finding a large number of seal impressions in the -
c - 9 I . '
lowermost of these floors. Given the extreme fragil-
ity of these artifacts and the difficulty in identifying Fig. 30. Mishrifeh. MB I1 figurine fragment.

'4 Excavations were funded by the National Endowment Study seasons were additionally supported by the Ahman-
for the Humanities, Ambassador International Cultural son Foundation. See reports in Orient-Express1995:3,67-70
Foundation, Neutrogena Corporation,and various donors. and AfO, in press.
134 HARVEY WEISS [AJA 101

Fig. 31. Mozan. Seal impression of Uqnitum, queen of Urkesh, found in Royal Store-
house AK, inv. 8a-2805, ca. 2300 B.C.

them (they are unbaked, very small, and for the most of Tell Brak. Our results depended upon the slow
part of the same color as the matrix within which and systematic recovery of the corpus of seal impres-
they are preserved), we proceeded with few work- sions and their painstaking interpretation. Out of
men and very close supervision. Each of the three the more than 600 seal impressions, over 170 of which
floor deposits was between 5 and 15 cm thick. We are inscribed, only six gave us the epigraphic clue
divided the lowermost floor into 1 x 1 m square loci we needed for the identification of the site.
and sifted the entire accumulation in each. We also The only other major artifacts found in this floor
took this opportunity to watersieve the entire floor deposit were figurines. More than 100 were found,
deposit for maximum recovery of botanical remains. not counting appendages and other fragments. An
Over the three seasons, we were able to complete exhaustive study has been undertaken by Rick
a descriptive, graphic, and photographic documen- Hauser, to whom we owe the following preliminary
tation of all the artifacts. description of the main types. Four types of human-
While we knew that the seal impressions were im- oid figurines have been distinguished according to
portant, we did not expect them to be so rich in in- how the artifact is terminated. Six body types have
formation. The most remarkable result obtained been established based on genus for animal figurines:
from a study of the seals was their identification of Bos,Ovis,Canis,Felis and other carnivores, Capra, and
the site as ancient Urkesh. But other results, too, Eguus.As might be expected, signs of domestication
were significant. A large number of different seals are readily apparent in figurines of dogs and sheep.
were evidently used by the same individual, in par- Curly-tailed dogs are present, with a heavy body type
ticular by the king and queen (fig. 31). An unexpect- similar to present-day chows or spitz. One dog figu-
edly close correlation exists between the legends and rine has not only this heavy body type but also a tail
the scenes, in other words, between the stated pro- carried high and a four-squarestance with fused but-
fession of the seal owner and the iconography rep- tocks, expressed sexual parts, and defined muscu-
resented on the seal. Finally, we have evidence for lature. The body type of the domesticated "fat-tailed"
an impressive new glyptic style, which is distinct from sheep is exemplified by another figurine: full torso,
that of southern Mesopotamia and even from that rounded rump, flat and rectangular tail, unexpressed
19971 ARCHAEOLOGY IN SYRIA 135
sexual parts, and little-expressedmusculature. Clear five areas explored so far. Three of these areas have
signs of domestication are present in the equid been continued or expanded from areas first inves-
figurines, such as depictions of the cheekstrap and tigated in 1983 and 1984.
browband of a harness. Many examples of Equus The greatest effort has been concentrated on the
figurines were found in third-millennium strata; Iron Age gateway in area A on the southern slope
further analysis may lead to differentiation by spe- of the prominent southern tell. The gate consists
cies within the genus. of a single 7.5 x 3.8 m room and a 7.5 x 3.2 m en-
The faunal remains were analyzed by Sandor trance porch. A five-stepstairway, located in the door-
Bokonyi, who was at Mozan for the duration of the way at the back of the main chamber, leads up to
1993 season, and he also worked closely with Hauser a 3-m-widestone-paved street that has been traced
to correlate faunal remains with animal figurines. more than 13 m to the north. The remains of occu-
MJe report here his major findings. The 7,601 identi- pation contemporary with the gate are preserved
fiable animal bones from Tell Mozan represent a at very different elevations ranging from about 0.3 m
faunal assemblage typical of a settlement. The pres- just inside the gate to more than 1.5m to the north,
ervation of the bones, however, is rather poor. The east, and west. The pottery associated with the gate-
caprovines are the most frequent animal species: way dates to the ninth and eighth centuries B.C. A
of these, 71% were identified as sheep and 29% as tower, preserved in seven courses of roughly shaped
goats. The next most frequent domesticated species blocks of basalt and limestone boulders, begins 80
is the pig, and then cattle, which represented only cm above the Iron Age street. The pottery in the lay-
a seventh of the bones. The remaining three domes. ers associated with and beneath the tower was very
tic species- horse, ass, and dog- have a much lower mixed, and Ayyubid pottery continues virtually down
frequency in the sample and therefore presumably to the gateway levels throughout the area.
a much lower importance in the economy of ancient A large scattering of stones to the northeast of the
Urkesh. Despite these low frequencies, however, the gateway represents collapse from its north wall.
presence of the domestic horse and donkey in the Persian-period pottery was associated with a tannur
MozanlUrkesh assemblage is exciting. Since neither and a rebuilding phase of this wall. Reuse into the
of the two species is related to domesticable wild Ayyubid period of the rebuilt and associated walls
stock in the area, their introduction to the north- obscures the evidence of Persian and Hellenistic oc.
western part of Southwest Asia implies trade con- cupation. Removal of the stone fall revealed a door-
nections as well as new possibilities for warfare and way through the gateway wall into what was appar-
mobility. Twenty-four wild species also occur, among ently a room to the east of the entryway to the gate.
them five ruminants and two ungulates (wild swine At least four Iron I1 building phases have been en-
and onager). This richness implies not only a large countered inside the gateway but are preserved only
amount of meat but also a potential for increasing to a height of half a meter.
the domestic stock from four domesticable wild East of the gate chamber a storeroom filled with
forms. Overall, however, bones of wild animals ac- broken storage vessels and charred beams is being
counted for only 2.4% of the total sample, while re- excavated but the floor contemporary with the gate-
mains of domesticated animals made up 97.6%, way has not yet been reached. A small cylinder seal
strongly suggesting a dominance of animal husband- was found near the top of the debris and large quan-
ry over hunting in the economic life of the site's third- tities of carbonized materials provided an excellent
millennium inhabitants. collection of botanical samples. The storage vessels
Qarqur. Rudolph H. Dornemann, American are similar in shape to a nearly complete wine jar
Schools of Oriental Research, reports: found with a collection of iron spear points in the
Excavations at Tell Qarqur were resumed with field destruction debris against the eastern wall of the
seasons from July to August in 1993, 1994, and main gate chamber.
1995.1Vell Qarqur is located 2 km east of the In our northernmost squares in area A, the Iron
Orontes River at the northern end of the Ghab. Ex- Age street was reached beneath 3 m of late fill, mostly
cavations have now reached Iron Age I1 levels in the deposited by erosion from higher areas of the tell.

l 3 Excavations were supported by the American Schools at the site, see J. Lundquist, "Tell Qarqur-The 1983 Sea-
of Oriental Research, the Catholic Biblical Association, son,'' AAS 33 (1983) 273-88.
and private contributions. For a report on the earlier work
136 HARVEY WEISS [AJA 101

The street was constructed by leveling heavily burnt tions of the area. In the center, particularly in square
and collapsed building remains and then raising the B2, the excavation continued systematically through
street to the desired height with earth and gravel the complex stratigraphy encountered in previous
fill. The heavily burnt bricks of a large building, with operations and then continued down to earlier lev-
walls at least 2 m thick and a partially stone-paved els. At least seven Iron I1 construction phases were
doorway that runs under the street, were being ar- evident. The earliest pottery encountered does not
ticulated as excavation concluded in 1995. Most of seem to date before the second half of the 10th cen-
the interior of the associated room in the northeast tury B.C. and most of the pottery dates to the ninth
of the area was removed by later pits. The function and eighth centuries. This assemblage includes typ-
and construction history of the building and its re- ical carinated bowls with a variety of surface treat-
lation to the foundations farther south are questions ments and large quantities of red-burnished platter
for future seasons. bowls, with reds that varied from browns and red-
The fill beneath the Iron Age street contained a browns early in the period to the strong reds of the
very heavy concentration of EB pottery, including eighth century. Materials characteristic of the sev-
a full range of 'Amuq G through J forms with an ex- enth and sixth centuries B.C. are scarce, and come
cellent representation of "Khirbet Kerak" and primarily from pits and surface contexts.
reserved-slip wares, and phase J painted and incised Area E, a trench on the northern slope of the south-
wares. A good representation of MB I1 sherds was ern tell, was opened in 1995. Though no stone walls
present, as well as several dark-burnished sherds from were encountered, stones of various sizes scattered
the Neolithic period, contemporary with 'Amuq across the area apparently represented tumble from
phases A and B. Incidental finds included Ubaid walls higher up the slope. A small exposure beneath
painted sherds, a triangular lug handle and beveled- the scatter of stones revealed mudbrick collapse,
rim bowls from the end of the fourth millennium burnt debris, and in situ brickwork at the tops of
B.C., and a small collection of Iron I sherds. One preserved and plastered walls. These remains were
of the most interesting objects was a seal impression associated with pottery that was primarily Early
on the outside rim of a storage jar dating to the Bronze IV in date, but included a few sherds of Iron
mid-third millennium. The simple design shows a Age I1 as well.
number of quadrupeds in superimposed rows and On the large, low northern tell, excavations re-
possibly a human figure facing the quadrupeds. The sumed in 1993 in three new squares adjacent to the
best parallels for this impression come from Tell area D excavations of 1983-1984. Though Ayyubid,
Mardikh. Late Byzantine, and Roman coins were found in the
As the large mass of stones exposed earlier be- area, the late occupation was poorly preserved. Foun-
neath the gateway were articulated in 1995, 10 steps dations of a Byzantine building and many ovens were
from a 3-m-widestairway were uncovered. The stair- encountered. Part of the area was disturbed by pit-
way was constructed against a retaining wall that ting, but good Iron Age stratigraphy was encountered
curved east, down and away from the line of the street along with the remains of stone and mudbrick walls
through the gateway. Again, the trench was intended of domestic buildings.
to explore Bronze Age remains below the gateway, The pottery associated with these levels is contem-
but we had only reached Iron Age levels by the end porary with the pottery from the gateway area and
of the season, and further down the slope excavation from the excavations in area B. A good number of
was suspended in Islamic levels. A trench dug for heavy storage vessels were found here but also well-
an irrigation pipeline 100 m south of the southern made vessels with painted decoration, the usual
tell encountered considerable Islamic building re- red-burnished wares, and imported Cypriot wares.
mains. It also documented Hellenistic, Persian, Iron The sample is too small to give a reliable indication
11, Iron I, MB 11, EB, Uruk, Ubaid, and Neolithic oc- of the function of the area at this point. If this area
cupation in the area. These remains support the as- is indicative of occupation on the low tell, the Iron
sertion that occupation at Qarqur extended south Age is much better represented than surface collec-
about 130 m beyond the edge of the high tell, as is tions and previous excavations led us to expect.
indicated by the contours of the area. We are just reaching the levels at which the con-
Excavations were also taken u p where the 1984 text of materials is secure enough stratigraphically
excavations left off in area B on the eastern edge that soil samples and bones will provide reliable ma-
of the high southern tell. Sections of mudbrick walls terials for botanical and zoological analysis. Initial
and foundations in many phases were badly cut by examination of selected samples from Iron I1 loci
pits, particularly in the northern and southern por. by palaeozoologist Susan Arter indicates that, despite
19971 ARCHAEOLOGY IN SYRIA 137

a heavy reliance on caprids, beef, and pork repre- On the eastern side, we traced room 2072 in which
sent nearly half of all the meat taken from domestic the tomb was constructed. Room 2072 has two open-
animals. The inhabitants of Qarqur were also ex- ings to the north, one to the slab court to the west,
ploiting wild resources including deer, bear, wild and one to the east leading to a small room (1235)
boar, turtles, birds, fish, and crabs. If the patterns less than 9 m*. Room 1235 contained 360 complete
of meat procurement and consumption now evident and fragmentary tablets. Some were found in niches
persist after the analysis of a much larger sample, in the southwest wall (one niche contained 40 tab-
they may reflect the food preferences of an adminis- lets), while others appeared to have fallen from
trative or diplomatic elite, or a:i urban population above.
with access to a wider variety of resources than is Excavation of the tomb and the room above it led
usually available to city dwellers, perhaps due in this to further excavation of the house and an attempt
case to the site's unique environmental setting. to place it in its architectural context. The tomb be-
Ras Shamra-Ugarit. Marguerite Yon, Mission longs to the category of large vaulted tombs of dressed
frangaise de Ras Shamra-Ougarit,Maison de ]'Orient, stone of the 13th century. The floor was covered with
Lyons, reports: large square slabs; funerary material of the last phase
The 1994 excavations, directed by M. Yon and Y. of inhumation included local, Mycenaean, and Cyp-
Calvet, were focused upon four areas (fig. 32). riot ceramics, and alabaster and faience objects.
"Centrede la ville." Two projects were undertaken: The archive discovered to date in the House of
1) soundings in house C to verify conclusions about Urtenu, about 600 tablets, comprises the largest ar-
the date of its construction, and to determine the chive retrieved at Ras Shamra outside of the Royal
eastern limits of street 1312; and 2) completion of Palace. Of the 1994 texts, 80% are in Akkadian, and
the excavation of house D. Two soundings confirmed others are in Ugaritic. Urtenu was apparently a con-
the construction of house C in the last phase of the siderable court personage in the last days of the king-
history of Ugarit, and established that the area be- dom. The historical significance of the archive prom-
tween house B and house D was an open space. In ises to be exceptional.
the mid-13th century the circulation route from 'Grand Rue." A new program of excavation in 1992
street 1038 to the north ended in an impasse, and in the southern sector defined the continuation of
to the south a narrow passage left street 35 and circulation route 3016, which went from the entrance
also terminated in an impasse. Then during the of the city to the center, following the riverside con-
final phase (perhaps at the beginning of the 12th structions and the large building B situated at the
century), the cross-walls and the large wall 130111309 northeast angle of the crossroads. Six weeks of ex-
were destroyed and removed; one had then a small cavation opened a 150-myarea. Excavation of build-
free passageway. But the constructions are poorly ing B, begun in 1994, is still in progress, but it is
preserved. already clear that its rooms are filled with storage
House of Urtenu. Excavation in 1986, 1988, and jars and masses of burnt materials, suggesting a vio.
1992 in the south-central zone, which in 1973 had lent fire.
produced an important series of texts, together with North Palace. Five levels of construction have been
subsequent stratigraphic study, defined the epi- identified in the street leading to the North Palace,
graphic archive and its context, the House of Urtenu dating from Middle Bronze IIB-C to Late Bronze
(fig. 33), named for its last owner, who lived at the I. The building situated under the North Palace is
beginning of the 12th century. Rectangular slabs contemporary with the House of the Ovens and was
found in 1992 seemed to be the cover of a tomb; perhaps constructed in Late Bronze I. The construc-
hence, the 1994 campaign had as its objective the tion of the North Palace has cut through this level.
definition of the archaeological context of the tomb The 1994 sounding therefore does not support a MB
and its excavation. date for the North Palace, which was constructed in
Extension of the excavation to the west allowed a sector already built up. The MB level was not de-
the investigation of new rooms. The house was aban- stroyed by the construction of the North Palace, but
doned at the beginning of the 12th century. But some by earlier LB I activity.
additions postdated the abandonment and could in- Sheikh HamadlDur-katlimmu. Hartmut Kiihne,
dicate a partial reoccupation. Material retrieved here Frei Universitat, Berlin, reports:
was of exceptional quality: alabaster vases, Mycenae- During the 13th-16th seasons, work at Tell Sheikh
an vessels, and utensils of stone and bronze. Fifty Hamad was concentrated in the central part of the
more tablets were also retrieved, apparently part of lower town. The original goal was to complete
the same archive of texts found earlier. the 1984-1987 excavation of the Neo-Assyrian resi-
 HARVEY WEISS [AJA 101

RAS SHAMRA - UGARIT: Excavated areas 1994

Fig. 32. Ras Shamra. Schematic plan of the site in 1994 and location of excavated areas.

Fig. 33. Ras Shamra. Schematic plan of the House of Urtenu, 1994.
 ARCHAEOLOGY IN SYRIA

dences of a type well known from the capital, As- grouped around a separate courtyard; the central
sur.l%t the end of 1992, a new building was discov- of three boxlike rooms on its eastern edge contained
ered east of the residence sector with its foundations one of the wells.
virtually cut into the former Assyrian residences. A The official part of the Red House was accessible
fire that had destroyed parts of the residences ap- from the north through an entrance hall and stair-
parently precipitated the founding of the new build- case. The central courtyard also allowed access to
ing, named the Red House because the walls of its the official part of the building via another staircase
rooms were covered with a red wash (fig. 34). While room in the northeastern corner. Two separate
larger parts of the residences were restored and courtyards had to be traversed toward the south to
reused, a destroyed room next to the new founda- reach a hall 16 m long that formed the main room
tion was simply covered with a new floor, thus seal- of the official part. This hall gave access to a number
ing the contents underneath. This room was filled of rooms in the south and in the southeast. To the
with inventory, thus marking the material culture east, the hall was connected with a small room and
at the time of destruction. a corridor, which led to a series of rooms in the east
In the room at the northwest corner of the new as well as the southeastern corner room of the build-
building (fig. 35), built directly into the walls of the ing. In this small room as well as in the corridor some
sealed inventory room, four unique cuneiform tab- 30 heart-shaped dockets were found in association
lets were discovered. They were written by an Assyr- with hundreds of jar sealings and fragmentary jars,
ian scribe, in the Assyrian language and according all of which had apparently fallen from the upper
to an Assyrian formula, but they were dated to the story on the occasion of a severe conflagration that
years 2 and 5 of the Babylonian king Nebuchadnez- destroyed the entire building. Some 200 jar sealings
zar 11. This seems to indicate that the foundation are stamp-sealed with about 20 different seals.
of the Red House took place in the posthssyrian The dockets deal with economic affairs; they are
period, implying that the destruction prompting the written in Aramaic and some are dated with Assyr-
new foundation was to be viewed in combination with ian eponyms. All of the recorded eponyms belong
the downfall of the Neo-Assyrian Empire. These tab- to the so-called"post-Canonic" period. This evidence
lets prove for the first time that the Assyrians and again raises the question of which Assyrian kings
the Assyrian way of life continued after the break- should be associated with these eponyms. In any case,
down of the empire and that the Assyrians did not they stem from the final years of the Assyrian Em-
disappear completely, as so many handbooks lead pire and were found in a context that, based on the
us to believe. typology of the building and on the evidence of
These results motivated a decision to excavate the above-mentioned tablets, is unhssyrian as well
the Red House extensively, if not completely, in the as post-Assyrian, a period rarely covered by material
course of the following seasons. By the end of the evidence.
1995 season, the ground plan of the whole building After the destruction of the building, some parts
was exposed. The building covers an area of 4,200 of it were reoccupied. In one room excavated in 1994,
m2, and was built with mudbrick walls averaging two ostraca were found that were dated on epigraph-
1.4 m in thickness. The courtyards are paved with ical grounds to the fifth century B.C., in other words,
bricks. Two wells furnished fresh water and a sophis- the time of the Achaemenid Empire just after the
ticated drainage system served all parts of the build- Late Babylonian period. This reoccupation seems
ing. Three staircases once led to an upper story. to close the history of lower town I1 in Dur-katlimmu,
The building apparently consists of a residential except for a Parthian-Roman cemetery that covered
part in the west and an official part in the east. The most of the area of the former Red House from the
entrance to the residential part is situated in the second century B.C. to the third century A.D.
south, consisting of a 14-m-long corridor leading Shiukh Fawqani. Luc Bachelot, CNRS, Paris, and
toward a large central courtyard. From this courtyard, Mario Fales, Universita degli Studi di Verona, report:
separate units are accessible, each composed of a O n the basis of the results obtained during the
large hall, a smaller hall, a bathroom, a kitchen, and first campaign at Shiukh Fawqani (October 1994),
one or more storerooms. The southwestern unit is the Expedition17decided to continue the excavation

'"or a report of previous excavations, see H. Kiihne, (GIRA) is formed of French, Italian, and Syrian archae-
"Tell Sheikh HamadlDur-katlimmu 1985-1987,"AfO 36-37 ologists, historians, and palaeoenvironmental specialists,
(1989-1990) 316-23. coordinated by Bachelot and Fales, and including Marta
Groupe international de recherches archCologiques Lucioani, Martin Makinson, and Daniele Morandi.
140 HARVEY WEISS [AJA 101

Fig. 34. Sheikh Hamad. Central courtyard and entrance corridor of the Red House, end of the seventh century B.C.

Fig. 35. Sheikh Hamad. Red House, archive room and corridor from the east.
19971 ARCHAEOLOGY IN SYRIA 141
in three of the five sectors that had been previously Sector 1;: Epigraphic discoveries of considerable
investigated: sectors D, E, and E importance were made beneath a group of Islamic
Sector D. Phase A, the most recent of the three tombs and eroded Roman levels represented by re-
phases identified, is represented by a group of struc- mains of terracing and some fragmentary floors and
tures in mudbrick, lying on foundations constructed walls. In a building and on the floor of a room, lo-
of large stones from the Euphrates riverbed. Erosion cated approximately 2 m beneath the present-day
has almost entirely destroyed the superstructures. summit of the tell, a small archive was found: it in.
A preliminary analysis of the pottery of this phase cluded two complete tablets and various fragments
suggests a date corresponding to EB I11 or IV. The inscribed in Aramaic alphabetic script, as well as As-
buildings of phase A occupy almost all the area of syrian texts. The largest of the Aramaic tablets bears
the two squares opened at present, and probably con- a legal text relating to a loan of silver. Three stamp-
tinued beyond these limits in all directions. seal impressions are placed on the upper margin
Phase B is characterized by what are by far the of the tablet; two of them would seem to represent
most impressive and best-preserved buildings discov. signs of the Neo-Hittite hieroglyphic repertoire. The
ered on the site this year, reaching to a height of second tablet from Tell Shiukh Fawqani appears to
approximately 3 m. Two rooms were entirely exca- represent the longest document dicovered so far
vated, while two more were traced on the surface. within the small corpus of Aramaic texts on clay tab-
A preliminary examination of the pottery dates this lets of Neo.Assyrian date. It is also a legal text. Both
architectural complex to EB I and 11 (first half of the Assyrian and the Aramaic texts are part of one
the third millennium). The jars are characteristic family archive, and demonstrate a general fluency
of this period, with everted rims and decorated neck in both languages and forms of writing during the
and shoulder. Also noteworthy is the lithic material, seventh century B.C. The small area cleared this year,
mainly "Canaanean blades" as well as blades of tri- in relation to these interesting finds, suggests that
angular section, some of considerable size (165 x future campaigns will allow us to learn more about
40 x 10 mm). The original floors of rooms 1 and everyday life and the history of the population in
2 and the limiting walls had been installed on the this remote province of the Assyrian Empire.
remains of previous buildings, which were taken Sector G. After villagers made a chance discovery
down, leveled, and used as foundations of the new of a bronze cup at the base of the eastern slope of
buildings. These remains form the main elements the tell, a small sounding (2 x 3 m) was excavated
of the phase C occupation. for 10 days. The first level consists of a tomb with
Sector E. Sector E was opened to continue recov- a complete jar. Just above its opening, we found a
ery of the portion of the settlement of the Late small overturned bronze bowl. It would seem that
Bronze Age that had yielded an exceptional pottery the burial pit was covered by mudbrick. In the sec-
assemblage (more than 30 complete forms), despite ond level, two further tombs were uncovered. One
the limitations of the sounding in 1994. The corner of them had held the bronze cup found by the vil-
of a room that we had not managed to clear in 1994 lagers. This burial pit was covered by mudbrick. The
was cleared during the second campaign. To the third level was only partially reached during the last
already rich pottery assemblage some 10 complete days of the campaign, and only a pise wall 0.55 m
forms have now been added, some of which seem thick was located. The concentration of these three
to represent previously unknown types. burials within an area not exceeding a few square
To extend the exploration of the destruction level meters suggests a complete necropolis in this sector.
here, an area of more than 50 m2 was opened Its excavation will be continued in the following
around the 1994 sounding. The top of the destruc- season.
tion level, corresponding to that of the room exca- Sweyhat. Richard L. Zettler and Michael D. Danti,
vated in the previous campaign, was only reached University of Pennsylvania, and Peter N. Peregrine,
at a depth of 3.80 m. A second room, measuring 5.30 Lawrence University, report:
x 6.50 m, was excavated to the east of the first one. The seventh and eighth seasons at Tell es-Sweyhat
Its characteristics are exactly the same: a floor ex. were conducted in the springs of 1993 and 1995.1*
tremely rich in material, sealed by a destruction layer Excavations on the west side of the main mound have
formed by walls and fallen roofing. now provided a sequence of eight major occupation

IHExcavations were sponsored by the University of manities and the National Geographic Society. For a re-
Pennsylvania Museum of Archaeology and Anthropology port on previous excavations, see Weiss (supra n. 11)139-42.
and supported by the National Endowment for the Hu-
142 HARVEY WEISS [AJA 101

phases spanning the third and early second millen- contiguous blocks of the lower town and to use the
nia B.C. Phase 4, which dates to the time of the remains we found on the ground to aid in interpret-
settlement's florescence in the late third millennium, ing the magnetic data. We collected data over two
includes the structure known as the "kitchen build- areas: block 8, consisting of 80 20 x 20 m grids on
ing," described in our report on the 1991 excavations. the site's western side, and block 9, 32 20 x 20 m
In preparation for the construction of the kitchen grids on the site's southern side. Analysis of the field
building and contemporary structures, the western data is currently in progress.
side of the existing mound was cut down ca. 0.95 m Our pilot program in 1993 tested the validity of
and a level surface was created. A stone retaining geomagnetic mapping by excavating an area (oper-
wall was then built against the cut face and raised ation 16) previously mapped at 1-m and 0.5-m inter-
to a level above the surface into which the cut was vals (fig. 36). Operation 16 yielded evidence of sev-
made. The area behind the retaining wall was filled eral building episodes, the best preserved of which
in, probably with soil dug out of the western side included parts of six loci that may have been exte-
of the mound. rior spaces. The most prominent features were
The kitchen building stood across an alleyway or horseshoe-shaped kilns. These yielded substantial
street from the area IV building and abutted the stone numbers of wasters and fused sherds, suggesting that
retaining wall. Remains uncovered to date include the kilns were used for firing pottery. Both the kilns
two rooms on each side of an L.shaped courtyard. and the lines of the walls showed up clearly in the
The kitchen building was rebuilt once in the course geomagnetic maps of the area.
of its existence. In addition to the kitchen building, In 1995, excavations were undertaken in two ad-
phase 4 included three rooms associated with two ditional areas mapped in 1993. Operation 23 was
buildings to the north. a 10 x 10 m square laid out over a marked geomag-
These phase 4 buildings existed on a lower level netic high, roughly circular in shape and about 5 m
than whatever stood to the east. The surface of the in diameter, and yielded a large circular kiln, ca. 5
fill behind the retaining wall revealed no trace of m in diameter, sitting in a shallow depression in vir-
buildings, and may have been open space. That open gin soil. Operation 25, a 2 x 30 m trench across the
space would have separated the kitchen building and eastern outer fortification wall, produced a sequence
the area IV industriallstorage facility on the west side of three major occupational phases. The second
of the main mound from structures at the center phase consisted of the outer fortification, built on
of the settlement. the leveled-off remains of the earlier buildings ten-
Earlier excavations suggested that the lower town tatively dated to the third quarter of the third mil-
represented an ideal field situation for mapping by lennium. The eastern wall, dated to the late third
remote sensing. In 1993, a pilot program was under- millennium, was apparently an earthen rampart
taken to determine the feasibility of geomagnetic faced on the outside with a sloping stone revetment
mapping. It employed a Geoscan FM-18gradiometer and on the inside with a mudbrick wall 20 m wide
provided by the University of Bradford's Department and set on stone footings. The outer fortification wall
of Archaeological Sciences and collected data using identified on the geomagnetic map of the area con-
a system of 400 data-pointgrids, with magnetic read- sisted of parallel linear low magnetic anomalies about
ings taken either at 1-m intervals over a 20 x 20 m 20 m wide. These excavations demonstrated that Tell
grid or at 0.5-m intervals over a 10 x 10 m grid. A es-Sweyhatwas already a substantial settlement cov-
total of 93 grids (86 large and 7 small) were collected ering the main mound and the eastern lower town
in seven separate areas (blocks 1-7) of the outer town. in the third quarter of the third millennium, and its
Our fieldwork in 1993, designed to test the reli- circumvallation was contemporary with the settle-
ability of the instrument and collection methods, ment's expansion in the late third millennium.
as well as the validity of the data for generating maps Planned excavations were interrupted in 1993
of archaeological features, proved the technique fea- when the landowner began irrigating the northwest-
sible. In 16 days, 4 ha of the lower town were mapped, ern lower town. The flow of water opened up several
and useful and informative archaeological maps sub- deep holes, revealing the location of shaft-and-
sequently created. Block I, for example, has provided chamber tombs. Salvage excavations were undertaken
a map showing the outer and inner fortification walls, on two open tombs, 001 and 002. The rectangular
a gateway in each wall, room blocks, a network of shaft of tomb 001, cut from the level of sterile soil,
streets, and an industrial area of kilns. was about 2 m deep. It ended on a narrow ledge, from
In 1995, the geomagnetic survey continued with which two steps led down to the floor of a chamber
two distinct aims: to collect additional data over large about 5 m in diameter. The 2.5-m-highchamber was
 ARCHAEOLOGY IN SYRIA

Fig. 36. Sweyhat. Plan of operation 16.

dome-shaped. Both tombs had been looted, but pot- The tomb contained at least 10 individuals, and also
tery, small finds, and bones confirmed their function yielded animal bones. The Tell es-Sweyhattombs are
and suggested a date in the third quarter of the third spread over an area of at least 1 ha on the north-
millennium. eastern periphery of the mid-third millennium vil-
A sinkhole in 1995 revealed the location of an un- lage, and the cemetery may include in all 100-150
looted tomb, and a 3 x 3 m unit was laid out and tombs.
excavated. The tomb chamber was oval, measuring Finally, in 1993 a new phase of regional work was
3.9 x 4.9 m. The floor contained two articulated skele- initiated to investigate-the evolution of the third-
tons, both on the western side of the chamber. The millennium settlement system surrounding Tell es-
smaller and more poorly preserved skeleton was asso- Sweyhat. Excavations have so far focused on Wilkin-
ciated with two crossed straight pins made of copper son's site 3, informally dubbed H a j i Ibrahim, which
o r bronze, limestone rings, and a series of beads at is 0.25 ha in size and rises 2.6 m above the plain.
the breast, including two small zoomorphic beads This site, located 500 m southeast of the outer for-
in the shape of a cow or bull. The arrangement of tification wall, is one of seven small early third-
pins and beads is similar to that worn by females millennium sites on wadis within 4 km of the Tell
shown on shell reliefs from Mari. Near the skeleton's es-Sweyhat.Aerial photographs show H a j i Ibrahim
head were two model wheels, perhaps originally asso- to have a distinct rectilinear form. Excavations have
ciated with a model wagon recovered from a large revealed two major phases of occupation dating to
pile of pottery against the north wall of the tomb the first half of the third millennium. These phases
chamber. The remainder of the human bones from produced at least four structures consisting of a sin-
the tomb were scattered around the chamber's floor. gle, central compound fronting on a courtyard sur-
144 HARVEY WEISS [AJA 101
rounded by one-roomed rectangular structures. One and Iron Age loci. A pre-Islamic destruction layer
of these mudbrick structures measured 4.6 x 3.8 m, was discovered during the first excavation season in
and material remains from associated contexts sug- 1987. Additional walls and floor surfaces associated
gest that it may have been a silo. Hajji Ibrahim was with this episode were investigated during 1995.Frag-
abandoned at the time of Tell es-Sweyhat'sexpansion ments of a blue-glazed Parthian-stylejug were found
in the late third millennium. in the destruction fill, as well as sherds from the shoul-
Tuneinir. Michael Fuller and Neathery Fuller, St. ders of unglazed jars decorated with the distinctive
Louis Community College, report: Parthian motif of an incised zigzag pattern. Sherds
The 1993,1994,and 1995 campaigns at Tell Tunei- from several large, tar-lined storagejars bear inscrip-
nir involved excavations in four sections of the tions written in tar with broad strokes; one of these
site.lg Research objectives included the definition consists of three letters written in the Parthian script
of Islamic-period commercial structures, identifica- typical of the second and third centuries A.D. Mud-
tion of features related to a church once used by a brick walls belonging to Late Roman houses were
Syriac-speakingcongregation, and the secure dating also uncovered during 1995. Associated artifacts in-
of various pre-Islamic strata. clude two fourth- or fifth-century lamps similar to
Area I. Horace Hummel, Concordia Seminary, examples from Dura Europos, and copper coins that
supervised the excavation of third-millennium loci confirm this dating.
in squares 61 and 62. A charcoal sample (Beta 59953) Church in area IZI. Excavation of a low mound
from these levels produced a radiocarbon date of along the eastern edge of Tuneinir, supervised by
2460-2030 B.C. (this and all other 14C dates given David Hanlon, St. Louis Community College, estab-
are calibrated values). The third-millennium loci are lished that a church existed at Tuneinir and was con-
predominantly ash-rich strata that contain moder- structed in four distinct architectural phases. In cur-
ate quantities of medium to coarse plainware, Me- rent seasons, our attention turned to the structure's
tallic ware, and Ninevite 5 sherds; clay figurine frag- periphery. A cistern was discovered in 1994 outside
ments representing sheep, goats, horses, and wagons; the northwest corner of the church. A number of
and flint sickle blades and bone tools. A clay token artifacts were found intentionally discarded in the
from this area corresponds to an astragalus-shaped cistern beneath 1 m of earth and rock fill. The most
type in Schmandt-Besserat's study of tokens, but is significant of these artifacts are two amber-colored
twice the size of similarly shaped tokens from Chag- glass oil lamps and a clear glass wine chalice (fig.
har Bazar and Nuzi. One of the few features discov- 37). A radiocarbon sample (Beta 74631) taken from
ered in these loci is a secondary burial of three frag- beneath the glass artifacts provided a date of 380-700
mentary human skulls in a pit lined with mudbrick A.D., while a sample from deeper in the cistern fill
and capped by basalt cobbles. A fault, possibly evi- (Beta 74632) was dated to 660-970 A.D. Byzantine,
dence of ancient earthquake activity, was also iden- Umayyad, and Abbasid artifacts such as lamps, pot.
tified in these squares. The fault, which extends into tery, and coins were found in several loci within the
the bedrock below, displaced beds an average of 30 cistern. Outside the cistern but also to the north-
cm and created several small cavities within the fill west of the church, a glass ampulla was found in 1995.
of the mound. The ampulla resembles a type of glass flask produced
Animal bones from third-millennium deposits, in the third century A.D. Ethnographic evidence sug-
analyzed by Michelle Loyet of Washington Univer- gests that the ampulla served as a container for
sity, consist almost entirely of domesticated animal chrism, a sacred oil used in anointing and in cere-
bones (sheep, goats, equids, and cattle) and a hand- monies of consecration.
ful of gazelle bones. Sheep and goats were by far the Market in area VI and khan in area VIZ. Sharon
most intensely utilized animals during the third mil- Doerre, University of Texas at Austin, supervised ex-
lennium. No remains of domesticated pigs have been cavation of Tuneinir's market. The 50-room ruin is
identified in the bone samples from area I. adjacent to the ruins of two khans and a bath. The
Excavations near the summit of the tell in area rectangular shops face the streets along the east and
I have uncovered several Islamic, Byzantine, Roman, north sides of the market. The remains suggest that

' T o r previous reports on the site, see M. Fuller and "A Medieval Church in Mesopotamia,"BA 57 (1994) 38-45;
N. Fuller, "Tell Tuneinir on the Khabur: Preliminary Re. "Vessels from an Early Christian Church," BA 57 (1994) 245;
port on Three Seasons," AAS 37-38 (198811989) 279-90; and "Continuity and Change in the Syriac Population at
"Tell Tuneinir 1986 and 1987,"AfO 36-37 (198911990)332-35. Tell Tuneinir, Syria," Arum 5 (forthcoming).
Reports concerning the church include Fuller and Fuller,
19971 ARCHAEOLOGY IN SYRIA 145

Fig. 37. Tuneinir. Glass wine chalice (center) and glass oil lamps from church in area 111.

meat and baked goods were sold from shops on the cm thick and were constructed with a core of mud-
north side, while shops along the east side sold jew- bricks faced with stone and mortar. This is an econ-
elry and fine pottery. The eroded plaster bases for omizing construction when compared with the solid
swifts, used in spinning and weaving, were recovered stone and mortar walls used in the south khan, bath,
in the upper fill of several shops, suggesting that tex- and market. Animal bones from the north khan con-
tile production occurred on the second floor above sist primarily of goat and sheep femur and humerus
nontextile shops, a pattern seen today in modern fragments. Copper coins from this area are Ayyubid-
Aleppo. The southwest corner of the market served period issues from the same mints identified in the
as a pottery yard and included a kiln, settling tank, market and south khan assemblages.
and well. Several hundred coins were recovered from Urnbashi. Frank Braemer, Institut francais d'ar-
the market during three seasons of excavation; the chhologie du Proche-Orient,Jean-Claude Echallier,
majority are Ayyubid, Zangid, and Urtukid dirhems CNRS, and Ahmed Taraqji, Directorate General of
and fils dating to the late 12th and 13th centuries. Antiquities, Damascus, report:
Copper coins minted at Sinjar, 90 km to the east, Two further field expeditions have allowed us to
are the most prevalent issue, but coins from the mints complete the exploration of Khirbet el-Umbashi,
of Mayafarakin, 200 km north, and Aleppo, 320 km which was started in 1991. We have refined the rel-
west, are also frequent. ative and absolute chronology of the site. The sites
A 2 x 4 m probe trench was excavated beneath of Hebariyeh and Khirbet ed-Dabab have also been
the 13th-century street surface in square 89 to test plotted, and the borders of West Safa and North Kraa
for earlier structures. A meter of mixed fill separated have been surveyed.
the Ayyubid market from an Abbasid cesspit filled The necropolises of Khirbet el-Umbashi include
with broken pottery, animal bones, and organic waste. at least 1,400 tombs of which 1,370 have been located.
Sterile soil was reached 4 m below street level in the Their period of use extends over more than four
deep probe. Bone collagen from the deepest levels centuries. The western area contains a previously
in the probe provided a radiocarbon date of 420-670 unknown type of tomb, consisting of a door and a
A.D. (Beta 74633). corridor for access to a burial vault.
Patrick Berger, St. Louis Community College, In the northern area of the site, some plans of
supervised the excavation of the north khan, the en- building remains belonging to the initial phase of
trance of which aligns with the north entrance of construction have been completed. Only the wall
the market in area VI. The dimensions of the north foundations have been preserved; these consist of
khan, 23.7 x 23.5 m, are almost identical to those a double face of small blocks on a ground-levelcourse.
of the original south khan before the latter structure The walls delimit small narrow cells with more or
was modified by the addition of a mosque, corral, less curved contours but not apse constructions
and shops. The walls of the north khan are 55-85 proper. House VN 1 01, which dates from the sec-
 HARVEY WEISS [AJA 101

ond and latest phase, has been wholly excavated; its diameter supported by peripheral posts without cen-
outer walls were furnished with benches, and the tral shoring. The second one comprises quadrangu-
roof was supported by a series of axial pillars. Cali- lar cells. To the south of the southern area, we have
brated radiocarbon dates have been obtained on the also located and cleaned structures marked by a low
earliest hearth level in this area (2837-2349 B.C.; OXA crescent.shaped wall and a flat area, suggesting the
LYON-65) and on ashes from the latest architec- remains of light camps.
tural phase of house VN 1 01 (2111-1789 B.C.; OXA Between Umbashi and Hebariyeh several differ-
LYON-64). ent types of water catchment and storage works were
In the zone between the northern and southern found, representing a sum of technological exper-
areas, in the western area, and, even further west- iments unparalleled in this period. The systems to
ward, on the slopes of the Umbashi valley, we have collect slope wash convey the water into the bed of
plotted a great number of rectangular megalithic the Wadi Umbashi, directly above the dam and cis-
constructions combined with one or several qua- tern connected with the rampart of the southern
drangular enclosures encompassed by a carefully area; they may have been designed to increase the
built stone wall. We consider such structures, at least flow into this cistern at the time when its walls were
provisionally, as a particular type of settlement in raised. The main cistern is fed by tapping the Wadi
a restricted chronological phase. This phase predates Umbashi above the western agglomeration of houses
the building of houses in the western area but post- by using a tapping dam, a wall that slants with re-
dates the first occupation of the site. spect to the wadi bed. A channel conveys the water
Within the area of the ossiferous deposit found down to a long natural basin, which had been ar-
previously in the southern area, we have discovered ranged to serve as a cistern. The overflow of this cis.
a lens of burnt earth in a layer between the rock and tern is discharged into two channels that carry the
the layer of bones. Precise redating of the bone de- excess to secondary cisterns close to the agglomera-
posit and, more importantly, of the ground level be- tion. A survey by electrical sounding of the main cis-
low this deposit and contemporary with the ram- tern allowed us to reconstruct the volume of water
part wall, is made possible by sampling: the date storage at over 30,000 m3. The dam connected with
obtained through an analysis carried out in Gronin- the southern rampart is the oldest such work pre-
gen in 1950 (4075f 160 B.P.) gives a calibrated date served in the Near East. It is an earthen dam, no
of 3020-2179 B.C., while the ground level contem- doubt based on a stone core set up directly on the
porary with the rampart is dated to 3331-2938 B.C. wadi bed.
(OXA LYON-63).The construction of the rampart, Surveys around Umbashi enabled us to locate a
and hence of its first hydraulic system, took place new site, Khirbet ed-Dabab, 7 km to the northeast
at the turning point between the fourth and third within the Safa massif. The site dates from the end
millennia. We have at last identified how the outline of the EB I1 period, and the agglomerated part has
of the l-ampart continues: at the summit of the south- been plotted in full. It is a nonfortified site; in the
western unit, we suspect that a 60 x 30 m citadel middle, we have noted a hierarchy within the struc-
was built onto the wall. If this hypothesis is correct, tures around the water source, which is next to a mon-
this building will be one of the oldest structures in umental tomb within an enclosure. On the periphery,
the southern area. clusters of houses are also separated by streets and
The western part of the site is composed of sev- lanes. The site includes some 50 houses preserved
eral units of heterogeneous ruins. The whole of this up to 3 m in height, showing construction techniques
agglomeration has been mapped at a scale of 1:500. quite different from those of Umbashi. Here, we do
Four sets of houses, separated by unbuilt areas, have not find roofs on pillars with slabs, but instead light
been singled out for study. Primarily EB IV ceramics roofs made of wood.
have been found in the cleared areas, but MB I1 ma- Plottingof another site, Hebariyeh, shows it to sup-
terial appeared during the excavation of the main port only seven half-buried houses, of the same type
cell of a house (VW 14) that revealed two floors with seen in the western part of Umbashi's southern area.
many grinding tools in place. It is thus a tiny village, but its houses are built upon
To the west of these agglomerated houses we have the remains of a large edifice, implying that, as at
plotted two camping areas. One dates to the Early Umbashi, occupation of this site may have originated
Bronze Age, while the other is poorly dated. Both from a fortified settlement. The hydraulic system ex-
camps lie along the channels that carry water from tends over several kilometers, diverting water from
the main southwestern cistern to the agglomeration. two small valleys into a large cistern.
The first camp includes round tents 2.7-3.5 m in Other than these two sites, no occupied site con-
19971 ARCHAEOLOGY IN SYRIA 147

temporaneous with Umbashi has been located. Khir- we can date the site from the ceramics to the EB 111111
bet Jerboua is a medieval fortified unit consisting transition; it was an aggregated, albeit nonfortified,
of several houses backed onto the lava flow. The west- settlement from the start.
ern border of the Safa, north of Umbashi, seems to Umm el-Marra. Hans H. Curvers, University of
be devoid of built-up settlement or even a concen- Amsterdam, and Glenn M. Schwartz,Johns Hopkins
tration of camps. In the central basin of the SafB, University, report:
no permanent habitation seems to exist. The only In 1994, the Johns Hopkins University and the
structures are the ever-present "desert kites:' which University of Amsterdam inaugurated a long-term
suggest, after study, a function linked to animal do- program of excavation and regional analysis at Tell
mestication and cattle control. A muddy basin (El Umm el-Marra in the Jabbul plain, east of Aleppo.
Aoui) situated on the western border of the Safa has The main aim of the project is to investigate the de-
revealed a large construction consisting of a 5-7 m velopmental trajectory of early complex society in
accumulation of stones that form small honeycomb the Aleppo region, a major core area of western Syr-
cells. The lithic material contains artifacts charac- ian urban civilization. Among the problems to be
teristic of the so-called "Khiamian" pre-Neolithic considered are the origins of complex society in the
period. It is the easternmost site in the survey area, third millennium B.C. (Early Bronze Age), the col-
located in the most desertic zone of this period. lapse at the end of the EB period, urban regenera-
This two-year effort has allowed us to reconstruct tion and formation of the Yambad state in the early
a general plan of the history of settlement in the second millennium (Middle Bronze Age), and the
Khirbet el-Umbashi area. An original pioneering set- incorporation of the area into the Mitannian and
tlement was established between 3300 and 2900 B.C. Hittite empires in the late second millennium (Late
and included a surrounding wall and a damming Bronze Age).
system. The area surrounded is about 5 ha, and no Umm el-Marra is the largest Bronze Age tell in
domestic building can be unequivocally attributed the Jabbul, with ca. 25 ha demarcated by a city wall.
to this phase. At Hebariyeh, settlement also origi- Soundings conducted between 1978 and 1982 by a
nates with a fortified structure. Belgian team directed by R. Tefnin identified Ro-
Subsequently, an extensive settlement system con- man and Hellenistic remains deposited above Early,
nected with pastoralism developed outside the walls, Middle, and Late Bronze occupation.20 Given these
consisting of at least four different building phases results, Matthiae suggested an identification of Umm
marked by quite different forms of construction: el-Marra with Tuba, a city mentioned frequently
megalithic structures with pillars, also present at in the Ebla archives that later became the capital
Hebariyeh; and a zone of hearths connected to of a vassal kingdom of Yambad. The Hopkinsl
smaller earthen structures built upon a foundation Amsterdam team has formulated a three-stage strat-
of small stones. In its maximum extension, settle- egy for the inauguration of its research in the Jab-
ment covered 40 ha near the beginning of the sec- bul. Stage 1, now completed, entailed two excavation
ond millennium B.C. seasons (May-June in both 1994 and 1995) to obtain
Lastly, a period of clustered settlement follows this broad exposure of second-millennium occupation
period of expansion and involves a minimum of two in a diversity of zones and to investigate the spatial
architectural phases, one consistingof scattered ham- organization of the Middle and Late Bronze city.
lets primarily in the western area and the other of Stage 2 will consist of a systematic surface survey
a village aggregated in the southern area occupied of the Jabbul plain, and in stage 3, we will aim for
until 1600-1400 B.C. The village, with about 30 a large excavated exposure of EB contexts at Umm
houses, covers ca. 2 ha. In this survey, the agglom- el-Marra.
eration of Khirbet ed-Daba appears as a counterpoint: Acropolis. Nine 10 x 10 m squares were excavated

"' Reports on this work include R. Tefnin, "Exploration mann ed., Halawa 1980-1986 (Bonn 1989) 19-56.
archeologique au nord du lac de Djabboul (sy;ie): Une Funding for the 1994-1995 excavations was provided
campagne de sondages sur le site d'Oumm el-Marra, 1978," by, among others, the National Endowment for the Hu-
Annwlire de 1 k t i t u t de philologie et d'histoire orientales et slaves manities, the National Geographic Society, and the Arthur
23 (1979) 71-94; and Tefnin, "Tall Umm al-Marra,"AfO 28 and Isadora Dellheim Foundation. See also the forthcom-
(198111982)235-39. For discussion of Tuba, see P. Matthiae, ing article by H.H. Curvers and G.M. Schwartz, "Umm el-
"Du-ub di Mardikh IIBl = Tu-ba di Alalakh VII," Studi Marra, a Bronze Age Urban Center in the Jabbul Plain,
Eblaiti 1 (1979) 115-18. The Halawa complex is published Western Syria," AJA 101:2 (1997).
by J. Meyer, "Die Grabungen in Planquadrat Q," in W. Orth-
148 H. WEISS, ARCHAEOLOGY IN SYRIA

in the western part of the site acropolis adjacent to this, a LB occupation included rectilinear domestic
the Belgian soundings where burned LB architecture architecture, a small kiln, and a street with steps lead-
(level IIb) had been identified. Minimal remains of ing up to the city wall area.
a Roman occupation (Ia) were stratified above large Transects. Given the scarcity of sherds on the site's
pits apparently contemporary with a thick-walled surface, systematic surface sherd collection was not
rectangular building of Hellenistic date (Ib). Prior undertaken. Instead, two transects were laid north-
to Ib were two LB phases. While the later phase (IIa) south and east-west across the site, and a 2 x 2 m
was poorly preserved, the earlier (IIb) included a pit was excavated every 20 m along these lines. Each
large complex of small rooms with homogenous do. square was taken down to its uppermost architectural
mestic debris. Given the uniform construction and feature. When ceramic analysis is concluded, a de-
contents of these rooms, we tentatively interpret them tailed picture of the temporal distribution of later
as centrally planned housing for dependent work- occupations across the site will be available.
ers, analogous to the MB complex excavated at Tell The MB phases exposed thus far contained ceram-
Halawa (Tell A, Planquadrat Q , level 2). Three phases ics characteristic of the middle and later segments
of MB (period 111) domestic occupation were iden. of the Syrian MB period. The LB corpus compares
tified below the period IIb architecture, all charac- with earlier LB (15th century B.C.) assemblages from
terized by intensive modification and reconstruction. Munbaqa (Ekalte)and Hadidi (Azu)on the Euphrates;
In the eastern acropolis area, results from the six particularly diagnostic were shallow bowls with in-
10 x 10 m squares excavated include burned LB ar- verted bead rims, as well as Nuzi ware and Cypriot
chitecture constructed above three MB phases, the White Slip I1 bowl sherds. Other aspects of LB ma.
earliest of which contained a set of rooms built terial culture also displayed similarities to that of
against a 2.6.m.thick mudbrick wall, perhaps an en- Emar (13th century B.C.), including clay humped-
closure wall for the acropolis. bull figurines and stone weights. The Hellenistic con-
West and northwest areas. Below Roman and Hel- texts were marked by "Persian" horse and rider
lenistic vestiges, the five 10 x 10 m squares excavated figurines, glazed gray oil lamps, and ring-based bowls
in west area A yielded evidence of burned LB do- with incurving rims and a dark wash, while Roman-
mestic architecture. The apparent contemporaneity period remains included oil lamps and other ceram-
of these remains with the burned level IIb occupa- ics characteristic of the later first century A.D.
tion on the acropolis is suggestive of a site.wide To conclude, the 1994 and 1995 excavations at
conflagration in the early LB period. The debris of Umm el-Marra have provided a foundation for fu.
one house included a small green-glazed ceramic ture research by documenting second-millennium,
jar and an Egyptian or Egyptianizing alabaster jar Hellenistic, and Roman occupation in a variety of
with duck.headed handles; a Mitannian Common- zones across the site. The extensive LB occupation
style faience cylinder seal was recovered from an out- was an unexpected discovery of potentially great sig-
side area in a square to the northeast. Found in the nificance, since the period is only modestly attested
same square, from deeper contexts, were the remains in the archaeology of inland Syria, and the scale and
of a large-scale, possibly public, structure of appar- character of the MB occupation also signals the site's
ent MB date. A 2-m-wide trench extending west to importance in the period of Yamhad. In addition
the tell edge revealed no evidence of a LB city wall to expanding and clarifying the second.millennium
but documented several MB phases of fortification; occupation, with attention to distribution patterns
strata sloping down to the east inside the MB wall in faunal, botanical, and ceramic data, plans for fu-
yielded a hematite cylinder seal of the style attrib- ture work include the initiation of a regional survey
uted by Collon to an "Aleppo workshop" associated project and the exposure of occupation remains from
with the Yambad royal court. East of west area A, the Early Bronze Age.
two 10 x 10 m squares were opened in west area
B, revealing a vast LB trash pit intruding into MB
and EB deposits resting on bedrock. DEPARTMENTS OF ANTHROPOLOGY AND
In addition, four 10 x 10 m squares were exca- NEAR EASTERN LANGUAGES AND CIVILIZATIONS
vated near the highest point of the tell in the north- YALE UNIVERSITY
west. Eroded Hellenistic architecture yielded a NEW HAVEN, CONNECTICUT 06520-8236
bronze ladle with a duck-headed handle, and below HARVEY .WEISSOYALE.EDU
http://www.jstor.org

LINKED CITATIONS
- Page 1 of 1 -

You have printed the following article:

Archaeology in Syria
Harvey Weiss
American Journal of Archaeology, Vol. 101, No. 1. (Jan., 1997), pp. 97-148.
Stable URL:
http://links.jstor.org/sici?sici=0002-9114%28199701%29101%3A1%3C97%3AAIS%3E2.0.CO%3B2-C

This article references the following linked citations. If you are trying to access articles from an
off-campus location, you may be required to first logon via your library web site to access JSTOR. Please
visit your library's website or contact a librarian to learn about options for remote access to JSTOR.

[Footnotes]

12
The Genesis and Collapse of Third Millennium North Mesopotamian Civilization
H. Weiss; M.-A. Courty; W. Wetterstrom; F. Guichard; L. Senior; R. Meadow; A. Curnow
Science, New Series, Vol. 261, No. 5124. (Aug. 20, 1993), pp. 995-1004.
Stable URL:
http://links.jstor.org/sici?sici=0036-8075%2819930820%293%3A261%3A5124%3C995%3ATGACOT%3E2.0.CO%3B2-4

19
A Medieval Church in Mesopotamia
Michael Fuller; Neathery Fuller
The Biblical Archaeologist, Vol. 57, No. 1. (Mar., 1994), pp. 38-45.
Stable URL:
http://links.jstor.org/sici?sici=0006-0895%28199403%2957%3A1%3C38%3AAMCIM%3E2.0.CO%3B2-M

20
Umm el-Marra, a Bronze Age Urban Center in the Jabbul Plain, Western Syria
Hans H. Curvers; Glenn M. Schwartz; Sally Dunham
American Journal of Archaeology, Vol. 101, No. 2. (Apr., 1997), pp. 201-239.
Stable URL:
http://links.jstor.org/sici?sici=0002-9114%28199704%29101%3A2%3C201%3AUEABAU%3E2.0.CO%3B2-S

NOTE: The reference numbering from the original has been maintained in this citation list.