archeology syria2

Archaeology in SyriaAuthor(s): Harvey WeissReviewed work(s):Source: American Journal of Archaeology, Vol. 98, No. 1 (Jan., 1994), pp. 101-158Published by: Archaeological Institute of AmericaStable URL: http://www.jstor.org/stable/506223 .Accessed: 07/02/2012 09:48

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

Archaeological Institute of America is collaborating with JSTOR to digitize, preserve and extend access toAmerican Journal of Archaeology.

http://www.jstor.org

http://www.jstor.org/action/showPublisher?publisherCode=aia

http://www.jstor.org/stable/506223?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp

Archaeology in Syria HARVEY WEISS

The pace of Syrian archaeological research continues to accelerate annually. Here we present 33 re-

ports, which reflect only half the number of foreign and Syrian active field research projects supervised by the Directorate General of Antiquities and Mu- seums, Damascus.' To document these efforts and the massive archaeological literature that they are

producing, we now have several vehicles in addition to the AJA survey. One is the recent prehistory survey of S. Muhesen.2 Another is the index and bibliography of Syrian archaeology prepared by M. Maqdissi, to be published shortly in Syria. A third is the annual

survey of excavations edited by the Directorate Gen- eral of Antiquities, Damascus, and published in AAS, and a fourth is the reborn "Archiologische Forschun-

gen in Syrien" to be published in AfO. Figure 1 illustrates the distribution of Syrian field

research. The Habur plains, associated Habur dam

salvage operations, the Euphrates River valley, and associated Tishrin dam salvage operations remain ma-

jor research foci. A majority of projects are oriented toward third-millennium, late prehistoric, and Palaeo- lithic problems, but large efforts are also focused

upon later periods as well. The largest foreign archaeological effort in Syria is

by French teams. Fifteen French or Syro-French excavations and six French survey projects were supported in 1992.3 Within the complex array of French

research, six Syrian research programs are supported by the Centre national de la recherche scientifique (CNRS), Unite de recherche associde 913 GREMO

(Groupe de recherche sur le Moyen-Orient), "De la mer au desert: Gestion de l'espace et organisation des

sociekts au Moyen-Orient," under the general direction of Remy Boucharlat. These diverse, often mul-

tidisciplinary programs extend beyond traditional limits of archaeological excavation and survey and deserve mention here.

The "Quaternaire et prehistoire" program, directed by P. Sanlaville, involves the collective research

ofJ. Besangon, L. Copeland, and S. Muhesen on the Orontes River and the Ghab Basin, including the Middle Acheulean site of Latamne. The results of this

project will soon be available in a three-part work entitled Peuplement et environnement au Paldolithique dans la valled de l'Oronte. Future plans of this group include additional work at Ard Hamed, near Tartous, an important Middle and Late Acheulean site, and a new project at Palmyra, part of a larger project fo-

cusing upon "les marges arides du Croissant Fertile," i.e., endoric basins, and extending to climatic studies of the Syrian steppe.

The program entitled "Le milieu et l'homme ia l'Holocene," directed by B. Geyer, is investigating the interaction between the environment and human societies, within two research projects. The first is the Lower Euphrates valley project, in coordination with

J.-C. Margueron at Mari, which focuses upon Neo- lithic to Islamic-period physical and human geogra- phy. The second, in coordination with G. Tate, is devoted to the "massif calcaire" of northern Syria and the natural forces that conditioned Byzantine settlement in this region.

A coastal research program directed by R. Dalonge- ville is reevaluating the Syrian evidence for late Hol- ocene tectonic movement and sedimentation, in addition to involving work along the Aegean, Persian Gulf, and Makran coasts. A sixth research enterprise, under the heading "Ethno-archaeology of the Near East," focuses upon the region of Qdeir-el-Kowm and

I Publication of this second newsletter on archaeology in Syria, as well as of the initial article in the series (AJA 95 [1991] 683-740), has been made possible by a generous subvention from the Ministry of Culture of the Syrian Arab

Republic and its Minister, Her Excellency Najah Attar. The editors of AJA and I are very grateful for this continuing support.

The contributors to this survey join me in thanking Ali Abu-Assaf, Director General of Antiquities and Museums, Damascus, and Adnan Bounni, Director of Excavations, for the collegial support extended to all projects, as well as the representatives of the Directorate General who have worked

so hard to insure the success of each project. The preparation of this survey of field research has benefited considerably from the editorial efforts of Indira Sweeny. Several contributions were translated by Sweeny, Ulla Kasten, and myself.

2 S. Muhesen, "Bilan de la prehistoire en Syrie," Syria 69 (1992) 247-74; Muhesen, "Bibliographie general sur la pr&- histoire de la Syrie," Syria 69 (1992) 275-304.

3 M. Al-Maqdissi, "Missioni regolari in Siria 1992," in O. Rouault and M.-G. Masetti-Rouault eds., L'Eufrate e il tempo (Rome 1993) 16-17.

101 American Journal of Archaeology 98 (1994) 101-58

102 HARVEY WEISS [AJA 98

TUR EY Qa

* 'Abr *Nustell

MembidiAKabir Quzaq 1W

aij

'Ain Dara Halula • Banat e Sabi AbyadM a Teinir Jebe Sinjar Qitar Jerf al-Ahmar *Kirbet esh-Shenef Jebel 'abd al-'Azz Mashnaqa

Sweyhat a Aleppo 4bd

iduib a * Afis Raqqa ,.

al-Bassit *7Ebla

ph. &Ma'arat en-Numan

* Resafa/Segiopolis Ras Ugarit Q Sheikh Hamad

0 * Apamea B\ it ama el

Kowm

Deir ez-Zor a

HaSYRIA SSalamiyah

*Kazel * Mislhrif i/Qatna

Homs Dura-Europos

0 Tadimor/Palmyra

LEBANON

SMoumnassakhin

A Damascus

*Sakka

SKh,

el-Umbashi

Suweida

-

JORDAN Der'a Mtoun ..

A J R A SBusra 0 50 t00

Kms.

- -I I I - I I

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

two villages, one occupied by sedentary agricultural- ists, the other occupied by seminomads (see el-Kowm, below).

Two recent historical studies should be mentioned for their exposure of the range of settlement and politico-economic relationships that have characterized the Euphrates and Habur regions. W. Hutterr6th and N. G6yuing are preparing an analysis of the 16th- century Jezireh landscape reconstructed from Otto- man tax records. A preliminary report of this project describes four features of interest to archaeologists: 1) rain-fed agriculture supporting a dense population

of villages with a distribution similar to that of today; 2) state-run rice plantations, controlled from Diyar- bakr, within the liwa of Nusaybin, which encompassed the territory between the Jagjagh and the Jarrah, exploiting forced labor from villages and tribes; 3) seminomadic tribesmen along the Habur River south of Hasseke engaged in spring/summer irrigation agriculture of millet and fall/winter steppe pastoralism, similar to the system described by Charles;4 and 4) few sedentary settlements and little cultivation along the Euphrates, which was then mostly occupied by pastoralists.5

H. Charles, Tribus moutonnieres du Moyen Euphrate (Damascus 1939).

5 W. Hiitteroth, "Villages and Tribes of the Gezira under

Early Ottoman Administration (16th Century): A Prelimi- nary Report," Berytus 38 (1990) 179-84.

1994] ARCHAEOLOGY IN SYRIA 103

Fig. 2. West Jezireh. Tell Hagib from the south.

A major study of the impact of the French mandate

upon the Jezireh has been concluded by C. Velud. Velud indicates that on the eve of the French occu-

pation of the Jezireh, the rain-fed agricultural regions of the Habur plains were virtually devoid of sedentary population but for small islands of agricultural settlement.6 Complementary data for the Jezireh are available in Bruinessen's recently translated study.7

Three major conferences were held in 1993, and we can anticipate the publication of significant con- ference volumes: 1) "Mari, Ebla et les Hourrites," sponsored by P. Garelli and J.-M. Durand, 28-29 May 1993, in Paris; 2) "La Djezire et l'Euphrate syrien de la protohistoire a la fin du second mill6naire av. J.C.: Tendances dans l'interpr6tation historique des don-

n6es nouvelles," sponsored by O. Rouault and M.

Wifler, 21-24 June 1993, in Paris; and 3) "Le pays d'Ougarit autour de 1200 av. J.-C.: Histoire et ar-

cheologie," sponsored by M. Sznycer, M. Yon, and P. Bordreuil, 28 June-1 July 1993, in Paris.

Recent books to appear include the following: P. Amiet, Sceaux et cylindres en hematite et pierres diverses (Ras Shamra-Ougarit IX, Corpus de cylindres de Ras Shamra II, Paris 1992); H. de Contenson, Prdhistoire de Ras Shamra (Ras Shamra-Ougarit VIII, Paris 1992); Y. Calvet and B. Geyer, Barrages antiques de Syrie (Collection de la Maison de l'Orient mediterraneen 21, serie archeologique 12, Lyons 1992); P. Courbin, Fouilles de Bassit: Tombes du Fer (Paris 1993); J.-M. Durand and J.-C. Margueron eds., MARI 7 (Paris 1993); H. Klengel, Syria 3000 to 300 B.C.: A Hand- book of Political History (Berlin 1992); F. Pinnock, Le

perle del Palazzo Reale G (Rome 1993); G. Tate, Les

campagnes de la Syrie du nord du IIf au VIIe siecle: Un

exemple d'expansion demographique et economique 'a la

fin de l'antiquiti 1 (BAHBeyrouth 133, Paris 1992); 0. Rouault and M.-G. Masetti-Rouault eds., L'Eufrate e il tempo (Rome 1993); S. Mazzoni ed., Tell Afis e l'etei del Ferro (Seminari di orientalistica 2, Pisa 1992); and S. Mazzoni, Le impronte su giara eblaite e siriane nel Bronzo antico (Materiali e studi archeologici di Ebla I, Rome 1992).

The site reports that follow were prepared by the excavators themselves. Save for one survey project, the reports are organized in chronological groups and, within each group, in alphabetical order by site.

ARCHAEOLOGICAL SURVEY

West Jezireh Survey. Berthold Einwag, Deutsches Archiologisches Institut, Damascus, reports:

In 1991 and 1992 the Deutsches Archiologisches Institut, Damascus, undertook an archaeological sur-

vey in the West Jezireh, the area enclosed by the

Euphrates, the Balih, and the Turkish border.8 Until now, no archaeological research had been carried out in this region (except for the 1928 French excavation in Arslan Tash), whereas the upper course of the

Euphrates in Syria, as well as the Balih valley, has been the subject of recent studies. Most of the research area has been affected by enormous changes resulting from various dam projects. In addition to this, rapid population growth and extensive land use are causing serious damage to archaeological sites. The fieldwork

began in the fall of 1991 and was continued during the spring and fall of 1992.

Some 150 sites were examined, covering the full

range of time from the Palaeolithic to Islamic medi- eval times. Most of the sites are located in the northern

6 C. Velud, "Regime des terres et structures agraires en Jezireh syrienne durant la premibre moitid du vingtibme siecle," in B. Cannon ed., Terroirs et socidtis au Maghreb et au Moyen Orient (Lyons 1987) 161-94; Velud, Une experience d'administration regionale en Syrie pendant le Mandatfrancais: Conquete, colonisation, et mise en valeur de la Jezire (Diss.

Lyons 1991). 7 M. van Bruinessen, Agha, Shaikh and State (London

1992). 8 B. Einwag, "Vorbericht iiber die archaologische Gelin-

debegehung in der Westgazira," DM 7 (in press).


part of the area, which is part of the dry farming belt. In particular, the plain of Sarug is rich in archaeological remains. Further to the south, settlements become rarer due to the drier conditions, and they are mainly lined up along wadis draining into the Euphrates and the Balih.

Tell Hagib (mentioned previously by Thureau-

Dangin) is the largest mound in the Sarug plain and was occupied from the Halaf to the Ayyubid period, including an Uruk-period settlement and a vast Early Bronze Age (EBA) occupation. The lower city is today covered by a modern village (fig. 2). Numerous Halaf sites of various sizes were uncovered, ranging from small villages to large plain settlements and high mounds. In the following Ubaid period, the settlements are fewer. A large number of EBA sites were also documented. One of these, Tell Kufaifa, 22 km southwest of Tell Abiad, was not resettled in later times and shows clearly the structures of the citadel and the surrounding circular city wall. Some stone-

settings are visible on the surface. First- and second- millennium sites are numerous in the northern part of the surveyed region. In the extreme south, most sites belong to later periods, particularly the Byzan- tine and Islamic periods. In Byzantine times, occu-

pation of the West Jezireh area covered not only the

plain, but also peripheral regions such as the limestone areas bordering the Sarug plain. Many early Abbasid settlements were found northwest of Raqqa; however, settlement traces later than the Ayyubid period are scarce.

PALAEOLITHIC-FOURTH MILLENNIUM B.C.

Habur Basin Project. Frank Hole, Yale University, reports:

The Habur Basin Project of Yale University is de-

signed to elucidate changes and variation in early agricultural economies from a series of villages in one of the heartlands of early civilization, northern Mes-

opotamia.9 Innovations in these economies-from the

development of new species of cereals and livestock, to the use of animal-drawn plows, to new storage techniques, food preparation, and the distribution and uses of these products-laid the foundations on

which the early civilizations were built. It is ironic that the oldest stages of agriculture are now well docu-

mented, but that the succeeding millennia of inno- vation and expansion across the Near East are known

only in the barest outline. In a departure from normal practice, which focuses

on the excavation and analysis of single sites and in which each project is staffed with different personnel and guided by different objectives, the Habur Basin

Project sampled 16 carefully selected sites that span more than 4,000 years and are situated in several distinct environmental zones in the middle stretch of the Habur River. The use of standardized methods for recovering and analyzing the botanical and faunal remains from each of these sites ensures a degree of

comparability of raw data and their subsequent inter-

pretation that has never before been achieved in the Near East. Using these materials, we expect to establish a baseline of data on village economies in a chang- ing environmental context that will enlarge the basis of understanding of all projects working in the area.

Although the primary focus was to recover charred

plant material and bones for the information they may hold for the interpretation of prehistoric envi- ronments and economies, we also collected ceramic and lithic artifacts, particularly from periods that have not been systematically or extensively excavated previously, and also a large number of radiocarbon sam-

ples. The artifacts are currently under analysis at Yale

University, and the botanical and zoological collections are being studied at the Smithsonian Institution

by Joy McCorriston and Melinda Zeder, respectively, and we expect to have a large series of dates within the year.

Some of the initial results include the discovery of sites that may document the actual migration of people from southern Mesopotamia into the Habur dur-

ing the mid-fifth and again in the late fourth millennium B.C. The first people brought pottery and burial practices typical of the southern sites, and may have attempted to introduce irrigation agriculture into this region. The second penetration is part of the

widespread Uruk expansion, but at two of our sites, Umm Qseir and Tell Kuran, the Uruk episodes are

9 We gratefully acknowledge the financial support of the National Science Foundation, the National Geographic So- ciety, the Peabody Museum, and the Coe Foundation of Yale University.

In Hasseke, members of the Taza family arranged for us to use a vehicle and an apartment and provided many other services and kindnesses. In 1992, Michel Fortin invited Joy McCorriston to recover samples from Tells Atij and Gudeda,

and Muntaha Sagieh provided plant samples from Tell Kerma for botanical analysis. The Project collaborates closely with the Danish team, headed by Ingolf Thuesen, working at Tell Mashnaqa (see below). The American field staff consisted of Frank Hole, Nicholas Kouchoukos, and Joy McCorriston of Yale University, Gregory A. Johnson, Hunter College, and Robert Brakenridge, Dartmouth Col-

lege.


very small, short-lived, and seem to have featured

hunting of onagers and gazelles. At Tell Kuran we recovered foot bones from 100 gazelles that evidently were killed in a single, massive hunt. According to Zeder's interpretation, the mix of wild and domestic animals exploited at each site varied according to

purely local considerations until the third millennium when hunting essentially ceased in favor of herds of domestic stock.

McCorriston, who has worked with both charred

plant remains and charcoal, sees some indications of

changes in the vegetation of the plain, such as the

presence of trees in the sixth millennium that do not occur in the area today, but the details of these

changes remain to be fully worked out. Her analysis also suggests that sites situated on the river may have been occupied year-round whereas those farther out on the plain may have been used only seasonally.

Recognizing that environmental changes may have been important in the history of settlement in this

region, we have begun geomorphological studies, fo-

cusing on the Habur itself. Robert Brakenridge has carried out some preliminary studies of flood deposits beneath and in the lower layers of sites and is cur-

rently using satellite imagery to develop a map that will become the basis for a geographical information

system (GIS). Halula. Miquel Molist, Universidad Aut6nomo de

Barcelona (UAB), reports: The site of Tell Halula is situated in northern Syria

on the right bank of the Euphrates near Jerablus.'o The large tell (360 x 300 m, 8 m ht.) was first inves-

tigated in 1986 by the mission to Tell el-Qitar from Melbourne University, and later in 1989 by M.C. Cau- vin (CNRS), Ahmet Taha (Palmyra Museum), and M. Molist (UAB). From the surveys and related surface

archaeological remains, the sequence of occupation seemed to extend from the Pre-Pottery Neolithic B

(PPNB) period to the Ubaid period. The objectives of the 1991 and 1992 archaeological

fieldwork were 1) to open several soundings in order to gather data concerning the vertical stratigraphic sequence of the site, and 2) to start intensive excavation related to the PPNB period.

The south and southeast PPNB sequence. Levels containing PPNB occupations are documented in the southern area of the tell, directly over virgin soil. The archaeological excavation in this area (sectors I-III), covering about 169 m2, allowed us to observe the stratigraphic sequence, which revealed a continuous occupation of the site. Occupation levels displayed architectural remains, with some phases indicating abandonment or leveling, as well as external occupation in areas represented by fireplaces with household refuse.

These excavations have allowed us to define a domestic rectangular architecture of multiroom type, made of mudbricks over stone foundations, with plastered floors and walls. Domestic units inside the houses include elevated ovens, fire pits, and silos. Evidence of burials has been found in two architectural areas. Mainly of infants and often partial, these burials were deposited in pits below the floors.

The archaeological remains related to this part of the tell include lithic implements: flint debris (flakes, blades, and bladelets), cores, and a large number of obsidian bladelets. Worked products are mainly represented by Byblos and oval-shaped arrowheads. Sic- kle-blade elements with traces of bitumen are also well attested, either by convex shoulder types or blades with inverse retouch. Scrapers and flake- or blade- burins were also found.

Among the miscellaneous objects found were decorated implements, such as beads made of a green stone, limestone, and obsidian. Clay animal figurines and grinding stones were also found.

All of the excavated strata from this area date to the PPNB period; variation in artifacts seems to indicate an evolution from the upper to the middle phases of this cultural period. The three '4C dates for the upper sequence of Tell Halula fall between 8500 and 7900 B.P.

In the southeast corner of the tell a new square was opened (sector VII) with an area of approximately 108 m2, revealing an occupation characterized by complex stone architecture, with walls preserved up to 1.3 m high. The currently excavated area shows a partial wall that seems to surround the area of occu-

10 The Spanish Mission Project in Tell Halula is made possible through its inclusion in the Spanish Archaeological Program Abroad of the Ministerio de cultura, through the Direccion general de bellas artes y archivos. We express our gratitude to Jose M. Luzon Nogue, Director General, Bellas artes y archivos, for his help and support, as well as our gratitude to the Embajada de Espafia and to the Centro

cultural hispanico in Damascus. Botanical and faunal analyses reported here were executed by D. Helmer (ERA 17 of CNRS, France) and M. Sana (UAB, Spain), G. Willcox (ERA 17 of CNRS, France) and M. Catala (UAB, Spain), and J. Anfruns (Laboratory of Anthropology, Biology Faculty, UAB).


pation, and on its interior side, a series of stone, oval-

rectangular structures, with 1.2-m wide walls, clearly differ from the domestic structures. This may be an enclosure wall, terrace, or other form of monumental architecture.

The archaeological remains from this sector consist

mostly of worked lithic products. The main tools include Amuq arrowheads, sickle-blades, and retouched blades. Other remains include bone implements and sherds from handmade pots; unfortunately, no com-

plete vessels were found. This architectural unit and its related occupation

levels seem, at the moment, to date to the end of the seventh or beginning of the eighth millennium B.C., or "final PPNB."

The upper part of the tell: The late occupations. Ex- cavation was undertaken to establish general stratigraphy in the upper part of the tell, consisting of a 50 m2 sounding, as well as various smaller ones (totaling 60 m2) in different areas of the tell. A stone wall, well

preserved at a depth of 2.2 m, was found and is associated with three pre-pottery levels. Initial construction of the wall took place in the second half of the fourth millennium. In the intermediate strata, the floors connected with the wall are associated with

pottery with vegetal inclusions, and Dark-Faced Bur- nished ware, indicating occupation in the first half of the fourth millennium. Finally, several strata reveal a

partial reconstruction of the wall, and gypsum and mudwash floors with sherds that could be related to a late Halaf period and transitional Ubaid period.

Palaeoeconomy and environmental data. In the PPNB levels, preliminary analysis suggests that wild animals such as cattle (Bos primigenius), equids, fallow deer, and gazelle were the main food sources. Sheep and

goats seem to be domesticated in the upper levels, and wild or semi-domesticated in the lower ones. There are also qualitative and quantitative variations that seem to indicate domestication of cattle in the

upper levels. Several different varieties of wheat have been iden-

tified at the site: Triticum aestivum/durum (free-thresh- ing bread wheat/hard wheat) and Triticum monococ- cum/dicoccum (einkorn wheat, emmer wheat), as well as remains that have been preliminarily identified as Quercus and Pistacia.

The archaeological excavations carried out at Tell Halula have confirmed the importance of northern Syria for our understanding of the Neolithic period. The site provides a continuous stratigraphic sequence

covering the transition between middle PPNB, upper PPNB, and ceramic Neolithic levels (identified as pre- Halaf), terminating in a more recent occupation dur-

ing the Halaf transitional/Ubaid period. Some of the more interesting aspects of the site are

the characteristics of the domestic architecture, burial

practices and ritual, and monumental constructions. The site is equally important for the study of the process of animal domestication, with the presence of the first domesticated goats in levels dated to middle

PPNB/upper PPNB, and the evidence for early domestication of cattle in the most recent levels (sector VII, final PPNB).

Jerablus-Tahtani. Edgar Peltenburg, University of

Edinburgh, reports: A survey of threatened sites in the northern zone

of the Tishrin dam area was conducted in March 1992 with the aim of locating sites with accessible Chalco- lithic-Bronze Age deposits. This was part of a broader research strategy designed to study interaction between resource-rich highland zones and urban production centers in the lowlands of the Near East. Of several suitable candidates, Jerablus-Tahtani com- mended itself by virtue of its plentiful Uruk ceramics, and trial excavations were undertaken there during September and October 1992.

The site of Jerablus-Tahtani is an oval mound, 180 x 220 m, rising some 16 m beside an active backwater on the right bank of the Euphrates River, some 3.5 km south of Jerablus. Woolley referred to it, Copeland and Moore surveyed it, and de Conten- son noted the existence of six periods of occupation."

In an attempt to balance rescue and research goals, our dual strategy is to establish the chronological scope of the site and expose Uruk deposits. To accom-

plish this, two soundings and two step trenches were

placed in squares 013 and P13 on the riverside edge of the mound where Uruk ceramics were eroding from its face. Mudbrick walls were also visible in the face to the north and south of these squares. The

deepest trench, P13, yielded an impressive 3.5 m of Uruk deposits without reaching virgin soil, but no

building levels were encountered.

Virgin soil lies below the alluvium that forms the

present agricultural land surface beside the tell. There is, therefore, little sign of a Holocene terrace, but frequent lenses of water-sorted pebbles in the lower levels of the eastern exposure indicate the dangers of locating any site on the banks of the Euphrates and, conversely, the overriding need of the occupants

" C.L. Woolley, Carchemish II: The Town Defences (Lon- don 1921) 38, pl. 2, fig. 5; P. Sanlaville ed., Holocene Settle-

ment in North Syria (BAR-IS 238, Oxford 1985) 53-70, fig. 14.


Fig. 3. Jerablus-Tahtani. Terracotta figurine.

to adopt such a high-risk strategy. Direct access to riverine transport, rather than agriculture, may thus have motivated settlers to situate their site here; geomorphological studies will be used to test this hypothesis.

Ecofacts. Issues of subsistence and economy have not received much attention in studies of the Uruk expansion, in spite of sweeping proposals that colo- nists imported most of their foodstuffs. Ecofactual evidence bears on the vexed issue of the nature of the

expansion. Faunal remains are well preserved and

preliminary identifications by P. Croft include cattle,

caprine, pig, dog, and equid (i.e., donkey). Carbon- ized flora (identification pending) occurred in discrete units and in beveled-rim bowls.

Chronology. Synchronisms between local and southern Mesopotamian sequences are of paramount im-

portance, but since the latter require better definition,

a range of 14C dates will be useful. Adequate stratified material is present for this purpose.

Recovered ceramics are largely contemporary with those at Habuba Kabira South. There are some indications of a longer sequence, like that at Tell Sheikh Hamad and Tell 'Abr, which point to a prolonged and more complex history for the Uruk expansion. Types include 1) conical cups with string-cut bases and dim-

pled rims, 2) bowls with thickened, inverted or flat rims, 3) flowerpots, and 4)jars with everted rims, nose

lugs, drop spouts, impressed cordons, or incised decoration. Red-Slip bowls, coarse trays, and Reserved-

Slip and Pseudo-Reserved-Slip wares occur infre-

quently (3% of sample); beveled-rim bowls are more common (30-40%). Also found were the rim of a stone bowl, a Canaanite flint blade, and indications of Bronze Age occupation, but these will be assessed in

subsequent seasons. Later occupation of the site is documented by the

discovery, i.a., of a moldmade terracotta figurine of the sixth-fourth centuries B.C. depicting a female

clasping a palmette below the breasts (fig. 3). Of special interest are the eagle and dove pendants (?) below the necklace, the fine modeling, and traces of red

paint in circles dangling from the elaborately decorated sash, on the left arm, and on the palmette branches.

Khirbet esh-Shenef. Peter M.M.G. Akkermans, National Museum of Antiquities, Leiden, reports:

Khirbet esh-Shenef is a small, inconspicuous mound, about 85 m in diameter and 1.7 m in height. The site is located in the Upper Balih valley of northern Syria, near the hamlet of Al-Atchana. Excavations at this site started in 1988 and were continued in 1991 as part of the investigations at nearby Tell Sabi

Abyad.12 It appeared that the topmost level of occu-

pation was of Middle Assyrian date (late 13th century B.C.) and most likely represented an isolated farm- stead, undoubtedly associated with the major Middle

Assyrian settlement at Tell Sabi Abyad. The lower strata of occupation all fit in the later Halaf period and date, according to a series of '4C dates, to around 4800 B.C. (uncalibrated).

Halafian Khirbet esh-Shenef was a small, short- lived settlement, hardly covering an area of 0.25 ha. Earlier it was reported that Khirbet esh-Shenef con-

12 K. Bartl, "Khirbet esh-Shenef, a Late Bronze Age Set- tlement in the Balih Valley, Northern Syria," Akkadica 67 (1990) 10-32; P.M.M.G. Akkermans and I. Rossmeisl, "Ex- cavations at Tell Sabi Abyad, Northern Syria: A Regional Center on the Assyrian Frontier," Akkadica 66 (1990) 13- 60; Akkermans, "Khirbet esh-Shenef," in H. Weiss, "Ar-

chaeology in Syria," AJA 95 (1991) 690-91; Akkermans and M. Teeuwisse, "Soundings at Khirbet esh-Shenef, a Later Halaf Site in the Balih Valley (Northern Syria)," AAS (in press); Akkermans and M. Le Mik're, "The 1988 Excavations at Tell Sabi Abyad, a Later Neolithic Village in Northern Syria," AJA 96 (1992) 1-22.


sisted of two Halaf occupation levels, the upper separated from the lower by an erosion layer. It now

appears that this apparent stratigraphic hiatus is absent: although there is a clear chronological distinc- tion perceptible between various structures, it seems that the site was continuously inhabited for a short

period of time. Whereas some structures went out of use and were demolished, others were renovated or

newly added. Particularly in the case of the very small

tholoi, it seems that these buildings served only for

very restricted periods of time and were, according to needs, rapidly replaced.

Circular mudbrick buildings varying in diameter between I and 5 m were predominant at Khirbet esh-

Shenef, in contrast with the Early Halaf site of Tell Sabi Abyad where the emphasis was on rectangular architecture. Until very recently it was felt that many of these Halaf hamlets and villages were built in a

haphazard manner, with the settlement layout very unstructured. Khirbet esh-Shenef, however, gave evidence of a well-defined village plan: most of the smaller tholoi used for storage, baking, etc., were concentrated on the western edge of the site, while circular or horseshoe-shaped ovens and hearths were located on the eastern side. The central part of the

village was occupied by a cluster of rather large circular and rectangular structures most likely used for

living. The tholoi in the center of the village were of

considerable size (ca. 5 m diam.) but all had remark-

ably thin walls, 25-30 cm in width. At least one of these buildings showed on the interior side a series of mudbrick buttresses placed at regular intervals. In

addition, this tholos had a small rectangular antechamber with rounded edges, which had been added at a later stage to the circular room. This antechamber

gave access to the building. A large rounded hearth lined with mudbricks of half size was found in the tholos. Along the outer facade of the building a small, shallow pit was found that contained three vessels covered by ashes. A small painted jar had been placed on the floor of the pit. The neck of the vessel had been broken off intentionally. In this jar a small, lightly burnished "milk jug" with spout was placed, in

its turn covered by a miniaturejar. A virtually identical

deposit was unearthed on the opposite side of the

building. Most likely these finds represent ritually buried offerings.

Next to this tholos a rectangular mudbrick building was unearthed, flanked by large tholoi. This very regularly built structure consisted of a series of small,

oblong rooms of almost identical size (2.8 x 1.2 m). In one of the rooms a round hearth was found. The exterior wall facade showed a series of buttresses and niches. This kind of construction seems to have been rooted in an Early Halaf building tradition, for a

rectangular building with a similar niched facade was found earlier at Tell Sabi Abyad, level 3.

In and around the various structures, considerable

quantities of finely painted ceramics were found as well as bone awls, flint implements, clay sling missiles,

clay spindle whorls, basalt grinders and mortars, and some fragments of large, baked clay, and painted figurines of both women and animals.

Sites like Khirbet esh-Shenef are commonly found in the Balih valley and seem to be characteristic of the final stage of the Halaf period in the region.

El-Kowm (Palaeolithic). Sultan Muhesen, Damas- cus University, reports:

The Palaeolithic occupation of the el-Kowm region has been under study since 1980, and at present approximately 200 sites have been discovered. Some of these have intact occupation floors with numerous

archaeological levels. Excavation and sondages have been executed at Nadaouiyeh I (excavation), Youal B

(sondage), and Qdeir 23 (sondages) (fig. 4). These sites were excavated by a Syro-Swiss team directed by S. Muhesen (Damascus University) and J.-M. Le Ten- sorer (Basel University). Umm el-Tlel was excavated

by a Syro-French team directed by S. Muhesen (Da- mascus University) and E. Boeda (CNRS). All are

open-air sites near ancient springs.13

Nadaouiyeh I. Nadaouiyeh I was discovered in 1980, and soundings were dug in 1985. Four systematic campaigns of excavation (1988-1992) have followed.

Five geological units were distinguished along a section 5 m deep. These are (from the deepest): unit

I: fine lacustral sediments containing three archaeo-

logical levels (5, 6, and 7); unit II: coarse sandy-stony sediments, partly eroded, with no clear archaeological levels; unit III: varved lacustral clay with carstic phenomena, archaeologically sterile; unit IV: clay-sandy sediments with a mixture of artifacts and ceramics,

containing a feature to be discussed below; and unit V: compact sediment of aeolian loess, containing an

important archaeological level.

13 J. Besanlon et al., "Le Paldolithique de l'El Kowm: Rapport prdliminaire," Paleorient 7:1 (1981) 33-35; F. Hours et al., "Premiers travaux sur le site acheulden de Nadaouiyeh I (El Kowm, Syrie)," Palkorient 9:2 (1983) 5-

13; S. Muhesen and E. Boeda, "Umm el Tlel: Rapport prdliminaire de la campagne de fouille 1991, 1992, rapport interne," unpublished report.


i I I I

Qdeir 23 ,

1Oum el-Tlal

. ..... Nadaouiyehl

?deir /

\..........

YouaI B

tSEBKHA 'I : .....ii•

' i /

... ..........

\• / 7':'L'

S//" Hummol 0?/ ( A

) //

\0:iiiii~l ,/ / e m.. Ssondage el - Kowm

Sexcavation f

\

W. Muqalb(

I I / 0 5km

Fig. 4. El-Kowm. Palaeolithic sites in the el-Kowm basin.

Seven archaeological levels were distinguished, all

belonging to the Lower Palaeolithic (Acheulean) period. All except level I were exposed vertically along a section with limited surface, and so further horizontal excavations are necessary to confirm their character. Archaeological levels are (from the earliest): level 7: the deepest, containing Middle/Upper Acheulean material; levels 6/5: containing Upper Acheulean material; level I: the latest level, horizontally excavated, containing homogeneous Final Acheulean material.

In addition, displaced Middle and Upper Palaeo- lithic material was retrieved. Important faunal remains and samples for absolute dating were collected. In conclusion, Nadaouiyeh I is at present the only known Syrian Acheulean site with multiple occupations. It is a key site for understanding the Acheulean

period in this region. Youal B. A sondage done in 1989 proved the exis-

tence of evolved Upper Acheulean material in a limited "pocket" along the Fataya valley.

Qdeir 23. A sondage was executed in 1991, and surface material proved the existence of an Acheulean site with at least two levels (Upper/Late Acheulean).

Umm el-Tlel. Palaeolithic levels were excavated in two campaigns, 1991-1992, and a complete sequence of occupation was revealed. Complex II, with Upper

Palaeolithic (Aurignacian), was the first Upper Pa- laeolithic material to be found in situ, other than at Yabroud III. Complex III provided a kind of transitional industry dating to Middle/Upper Palaeolithic.

Complexes IV and V were Middle Palaeolithic, with a Levantine Mousterian industry. Further excavations will concentrate mainly on these Mousterian levels. Traces of hearths and rich faunal remains with indications of butchery were recovered. Samples for de-

termining climatic conditions and absolute dates were taken. Excavation here did not reach sterile bedrock, and transitional Lower/Middle Palaeolithic industries are expected to be found, as was indicated by surface artifacts belonging to the Hummalian and Yabrou- dian cultures.

Surveying for raw materials began in 1990 and continued during 1991. Many new Palaeolithic sites were discovered, as well as flint quarries and work-

shops, mainly in the mountains surrounding the basin of el-Kowm (fig. 4).

El-Kowm. J. Cauvin, Director, and D. Stordeur, Assistant Director, Permanent Mission at el-Kowm, report:

The basin of el-Kowm (15 km diam.) contains an unusual succession of archaeological occupations. The permanent French mission has been based here


since 1978. Besides several survey programs that led to the location of some 100 sites, and a systematic sampling operation on one of these, several excavations have been carried out. Three of these excavations were financed by the Direction g6nerale de recherches (DGRCST) of the Ministere des affaires

6trangeres: el-Kowm 2-Caracol, Qdeir 1, and Umm el-Tlel (with the participation of the UAB, Spain). The fourth, Nadaouiyeh I, was financed by the Uni-

versity of Basel, Switzerland.'4 The site of Nadaouiyeh I contains mainly occupa-

tions of the Lower Palaeolithic (see above). El-Kowm

2-Caracol, excavated between 1978 and 1986 by D. Stordeur, is in the process of being published. Mainly occupied by a succession of villages of late PPNB between 8000 and 7600 B.P., it has also produced Neolithic and Late Uruk levels. Umm el-Tlel and

Qdeir 1 are currently being excavated. This summary concerns the work carried out in 1991 and 1992 at these two sites.

Qdeir 1. This site, purely late PPNB (sometimes called "PPNC"), was first sampled systematically by J. Cauvin in 1979 and was then the object of a sounding by O. Aurenche in 1980. We returned to this site in 1989 (D. Stordeur) and have carried out two seasons of excavation.

The reason for opening the excavation again was

primarily to test the hypothesis put forward by J. Cauvin in light of the first data recovered: Was Qdeir 1 a base camp for pastoral nomads of the late PPNB, as the near absence of architectural construction seems to indicate? And if this were the case, was the site contemporary with the important sedentary vil-

lage occupation of el-Kowm, situated 8 km away? This

hypothesis also generated ethnological observation of the two sites today. The present village of el-Kowm is the only important sedentary settlement in the basin while Qdeir has long remained a seasonal camp for nomads now in the process of sedentarization. Our work thus has an ethnoarchaeological viewpoint: test-

ing for nomadic phenomena. The first results are stratigraphic in nature. Sound-

ings spread over the entire site have revealed two

types of successive occupation. The first, represented

by fine layers with objects strictly in situ, reveals no trace of architecture. The second, which follows without rupture, contains several occupation levels where

invariably there coexist a lone construction and large exterior expanses wherein artifactual material is abundant but in secondary deposits.

To these results may be added those produced by the extensive excavation of the earliest levels. In these we find workshop areas for flint-knappingjuxtaposed with habitation areas. The habitation areas were not excavated over a wide enough surface to reveal the

organization of the remains. The structures that have been recovered are for the moment limited to a few undefined hearths and a few faunal groups and stone

alignments. The objects consist of plaster vessels, which sometimes have basket imprints, stone vessels

(perhaps imported from the Euphrates), and a lithic

industry typical of the late desert PPNB (M.-C. Cau-

vin). Tools with traces of gloss indicate the practice of

agriculture from the earliest levels (microwear anal-

ysis carried out by P. Anderson). The faunal remains

(studied by D. Helmer) reveal the herding of sheep and a few goats.

No radical changes are noted, either in the objects or in the food remains, from the beginning to the end of occupation. Only the mode of settlement (at first

without, then with constructed houses) seems to vary distinctly. On the other hand, the comparisons established with the site of el-Kowm 2 seem to indicate differences of custom (technologies, stock management, etc.) and probably geographic origin as well for the inhabitants of the two sites.

We believe we have brought together several ar-

guments in favor of the seminomadic status of Qdeir (absence of construction, followed by construction; characteristics of the artifacts; and other elements). If our results continue to point in this direction, Qdeir could be one of the first sites excavated where semi- nomadism is clearly attested for the Neolithic.

Umm el-Tlel. This site presents an unusual strati-

graphical sequence. Roman levels crown the site and cover an occupation of late PPNB, similar to el-Kowm, but relating rather to seminomadic features such as are found at Qdeir.

" J. Cauvin, "Cinq annees de recherche (1978-1983) dans l'oasis d'El Kowm (Syrie)," AAS 33:1 (1983) 165-77; J. Cau- vin, "L'occupation prehistorique du desert syrien: Nouvelles recherches dans la cuvette d'El Kowm (1984-1989)," AAS 37-38 (1987-1988) 51-65; M.-C. Cauvin and E. Coqueug- niot, Techniques d'echantillonnage et analyse spatiale: Le campement epipaldolithique de Nadaouiyeh 2 (El Kowm, Syrie) (BAR-IS 522, Oxford 1989); 0. Aurenche and M.-C. Cauvin, "Qdeir 1, campagne 1980. Une installation neolithique du

VIIe millenaire," Cahiers de l'Euphrate 3 (1982) 51-77; J. Cauvin, "Les origines prehistoriques du nomadisme pastoral dans les pays du Levant: Le cas de l'Oasis d'El Kowm (Syrie)," in H.-P. Francfort ed., Nomades et Sidentaires en Asie centrale. Colloque franco-sovietique Alma-Ata, 1987 (Paris 1990) 69- 80; 0. Aurenche, Ethnoarchiologie du Proche-Orient: Rapport 1981 (Lyons 1981); M. Molist and M.-C. Cauvin, "Une nouvelle sequence stratifiee pour la prehistoire en Syrie deser- tique," Pallorient 16 (1990) 55-63.


Under the Neolithic occupation, important levels of the Epipalaeolithic (M. Molist and G. Alcalde of the UAB, and M.-C. Cauvin, CNRS) and the Palaeo- lithic (S. Muhesen, Damascus University, and E. Boeda, CNRS, see above) succeed each other.

All the levels of Geometric Kebaran are interstratified in a dune system. In 1992, a layer with a high content of carbonized remains was excavated, uncov-

ering over 40 m'. The lithic industry consists mostly of geometric microliths (rectangles) typical of this

period. The fauna is varied, with equids, gazelle, wolf, and camel (D. Helmer, CNRS, and M. Sana, UAB).

Sixteen levels have already been identified for the

Upper Palaeolithic, rarely found in situ in the Near East. The lithic industry tends clearly toward microliths (microbladelets). The bone projectile points are reminiscent of the Aurignacian points of Ksar Aqil, but it has not yet been demonstrated that these levels

belong to the Aurignacian of the Levant (present elsewhere in the basin).

Several layers are interspersed between the levels of the Upper and Middle Palaeolithic (see below). Those that follow the Mousterian have a blade indus-

try with tools reminiscent of the European Upper Palaeolithic or Ahmarian (Sinai). Curiously, these industries do not give the impression of evolving, either in relation to the preceding Mousterian or to the

Upper Palaeolithic that follows. A chronostratigraphical survey, in the form of ref-

erence soundings, has revealed 15 Mousterian ar-

chaeological layers of which some, better preserved, were selected to extend the excavation. Soundings were also dug at intervals over 60 m to explore the extent of the occupation. We have determined that the site was a large settlement by a lake or a marsh

(suggested by peaty layers interstratified with the ar-

chaeological layers). Thus, we are dealing here with

widely extended human occupations on the edge of a landlocked basin with a spring-fed lake.

The lithic industry belongs to the great Near East- ern Mousterian family. A careful study of the stratigraphy has allowed us to identify distinct technological traditions. The analysis of the tools has led to the detection of different activities in the same level.

Each of the Mousterian levels excavated in 1992 contains rich faunal remains (identified by Christophe

Griggo), well preserved and localized. The animals identified are horse, donkey, gazelle, rhinoceros, au- roch, deer, wolf, camel, and a few bones from birds and microfauna.

The bone assemblages have characteristic traces associated with flint artifacts that have become worn and then been resharpened. Such traces allow us to reconstruct customs related to hunting and to the treatment of the game. A reconstruction on a large scale of these activities may be envisioned for the future.

Even if the examples are few (six objects spread over several layers), one may speak of the existence of a Mousterian bone industry at Umm el-Tlel; artifacts that were clearly worked and used have been found. This material should be examined with care as it represents a technology still little understood for these periods.

The stratigraphy of Umm el-Tlel is now known to 3 m below the present soil surface and may be roughly correlated to climatic periods. The Kebaran Geomet- ric, the Neolithic, and the Roman period seem to correspond to arid phases, in contrast to the earlier periods. The Upper Palaeolithic occupation could have had a spring nearby. The Middle Palaeolithic seems to correspond to a humid period in association with a lake. A more precise climatic study of the entire basin of el-Kowm is underway.

Mashnaqa. Ingolf Thuesen, Carsten Niebuhr In- stitute, University of Copenhagen, reports:

In the early fall of 1991, a Danish archaeological expedition to the Habur continued the sampling of Ubaid remains at Tell Mashnaqa in the Middle Habur region. The main purpose was to expand the amount of information gained from a previous season of excavation (1990).'15 During five weeks of intensive excavation we exposed approximately 70 m2 of architectural remains from the first settlement stratum and the remains of four burials. The total exposure of the Ubaid settlement is now approximately 160 m2.

The stratigraphy observed in the 1990 season was confirmed. When cutting deeper into the main tell formation, however, three additional strata were encountered from settlements following the Ubaid period (IV-VI). The stratigraphy after the 1991 season therefore includes six strata (I-VI).

15 The expedition was made possible by contributions from the Carlsberg Foundation, Novo's Foundation, Knud Hojgaards Foundation, the Danish Research Council for the Humanities, and Generalkonsul Gosta Enboms Foundation. For a preliminary report on the 1990 season, see Weiss (supra n. 12) 691-92. DNA extractions in situ were carried

out by Henrik Nielsen, University of Copenhagen; dental analyses were carried out by Jens Gejlager and Lene R. Nielsen, University of Copenhagen; XRF analyses were made by Gerwulf Schneider, Freie Universitat, Berlin; faunal remains were examined by Melinda Zeder, and botanical remains by Joy McCorriston.


-;Y

' ' '' ''~' , ?:?_?~??.?:: : .:~:`:~i':::::i':.i:~ ?:?:?::?:?:?::::?:?:?: :~~~~~:~~ '' :?r?..?:?:?'?.. .. 1.

:?:?~;~:?:?';':?:?: =??. I' ~?:?:?~.?:?:??.?. ... .? ;?r .r ~? ("':' ~.???`.

.i' ... Y.. ::?:;:i:::?:::::::: -.c;e ,I s~. ?,, r. :~:?:?ti:?:?I?::::

i? ::i:::::i~::::::: :I~?:? "~~:~ ;~??" ":/'.? :?::::::?I::j:::::;:.' ''`~'~~''' :ij:::~i':::~::'::::~:: i , I :?::::i:::?j:::~:::::: ?. ....

o ''~'~'''`'`' :::j~f::i:':~::~~::: ::::?i'::'::?:::::f:=:.:i~: ~?'?~;???~?'; :::?:i:::::::

?'?' -?~? ??~?'?? '-'

-y Fig. 5. Mashnaqa. Fragment of terracotta plate or lid with stamp seal impression.

The main target for excavation was stratum I, the earliest evidence at the site of the Ubaid expansion from southern Mesopotamia to the northwest along the Euphrates-Habur river system. The 1991 excavation exposed more of the building complex found

during the previous season. This building had at least two parallel long rooms: the eastern wall of a long room showed rounded buttresses on the exterior face, and to the north, the long rooms are associated with smaller rooms.

There are no remains of architecture from stratum

II. The stratum consists of a thick layer of greenish- gray soil, into which burials have been dug. Stratum III contains non-domestic structures, e.g., a large mudbrick kiln and a platform. Strata IV-VI were only exposed in a limited area, but have yielded a pottery sequence and '4C samples.

The burials were found in the area to the north of the building complex both in stratum I (contemporary with the building) and in stratum II. One badly preserved burial may belong to stratum III. The burials are similar, consisting of a burial pit in which a wall has been built, so that the top courses of the wall would have been visible after the covering of the body. The body rests on the side along the wall. As a rule, the burial is furnished with one or more pottery vessels and perhaps a few smaller objects (sling bullets, spindle whorls, bone awls, etc.). A jar dug down beneath the floor within the stratum I building may have contained an infant burial; however, the jar did not contain visible remains of a skeleton. In the 1991 season intensive dental studies and DNA analyses were carried out in the field.

Main categories of finds include pottery and

chipped stone tools. The pottery sequence shows a

possible gap between strata III and IV. The former is characterized by a high index (>25% rim sherds) of painted vessels, while the latter is almost lacking painted pottery. The painted decoration falls within the Ubaid 3 tradition, and places the settlement in an

early phase of the Ubaid expansion into the Habur and Syria.

The chipped stone tools comprise more than 1,600 worked pieces of which more than 20% are made of obsidian. The obsidian can be traced to the Nemrut

Dag sources, according to elemental composition. Remarkable finds include two fragments of boat

models. The models are made from terracotta and show a long flat-bottomed type of boat, well suited for river navigation. The exterior face of the boats shows

painted decorations in typical Ubaid style. Among the sherds a stamp seal impression was identified, placed on a circular handmade plate or lid of terracotta (fig. 5). The geometric design has parallels in motifs on

stamp seals found, for example, at Tepe Gawra. Other finds include ground stone tools, bone tools, terracotta animal figurines, and beads.

All the above-mentioned finds are presently being analyzed in laboratories in Copenhagen, and the faunal and floral remains are being studied in the Smithsonian Institution. A series of carbon samples have been submitted to the AMS Laboratory, Univer-

sity of Aarhus. The preliminary results confirm the

chronological position of Ubaid settlement in the late sixth millennium B.C. It is our intention to publish a

comprehensive report on the 1990 and 1991 seasons

during 1993. Nustell. Helga Seeden, American University of

Beirut (AUB), and Ursula Wittwer-Backofen, An-

thropological Institute, Giessen University, report:


Fig. 6. Nustell. Jar burial 5, a 4-6 month-old infant.

The 1991 discovery and rescue of material from a Phoenician children's cemetery at Tyre in Lebanon

prompted a more detailed investigation into the infant burials discovered at Nustell in the Wadi Rijle area northwest of Hasseke.16 Such burials and recent cemeteries reserved for children have been found in

Syria and Lebanon, indicating that separate burial

grounds and practices for infants may have a very long tradition in the Near East.

During the AUB rescue excavations at Nustell, 12 architectural phases of successive mudbrick houses with associated domestic and communal activity areas were brought to light, dating to the late fourth and third millennia B.C. The 2-m high walls of the last and most substantial mudbrick house reached the

present surface of the tell. Within the 43 strata of the 12 major building phases, eight burials were found inside the excavated part of the village settlement. The only adult, a woman, was buried in an abandoned silo, which had been reused as a pit for ashes, brick

fragments, broken pottery, and other domestic debris. A second silo, near the first, contained another burial, of an infant three to four years old. The woman and the child had been buried in a flexed position. Both silos belong to the earliest settlement on the site,

consisting of small-room brick dwellings, some walls of hand-molded bricks, and silos dug into the natural

wadi gravel. The silos and the nearby house belong to the first, Uruk-period, occupation. During this period, the two excavated silos were used for burials, after they had been half-filled with occupational debris. Between the later mudbrick houses, which suc- ceeded those of the earliest settlers at the site, another seven child burials were discovered, five of them care-

fully placed in storage jars or other household pots, some closed with either a large bowl (fig. 6), sherd, or brick. The jars that were found undisturbed, or at most crushed, contained infants placed in fetal position and carefully buried near the doorsill of a house, under a courtyard floor, or, in one occasion, in a disused oven in a courtyard corner.

The anthropologist's report states that of the eight skeletons available for study, seven were children of less than four years of age. Most of them died at birth or survived only a few months. All the infants suffered from severe nutritional stress, which manifested itself in the bones already during pregnancy. Only one

three-year-old child, the oldest buried at Nustell and found in the silo, did not show these pathological changes. As an effect of this, the newborns were not

strong enough to survive any stress factors. These

pathologies may not be interpreted as the direct cause of death, but they weakened the infants. The only adult skeleton belonged to an aged woman, who suffered from skeletal degeneration.

All these infants therefore died at a very early age, or were in fact stillborn. Special care was taken to

bury them, probably at once, within the immediate home context, most likely by the parents. The only adult buried in the silo of the earliest settlement was a woman who had reached old age. Adults were not

usually buried inside the villages. The presence of such burials on a site probably indicates that the settlement has been abandoned and become a cemetery for a neighboring village. By contrast, child burials inside settlements are quite common and occurred at all periods. Numerous children were found buried, either in jars or in mudbrick graves, at the neighboring Bronze Age site of Tell Abu Hgaira. The evidence indicates that the victims of infant mortality at Nustell in the third and fourth millennia were buried with

special care, in closed storage jars or other domestic vessels.

Scarcely a household at Nustell had not lost one or more of its children in the first five years of life, a phenomenon observed as late as this century (1938) in a Lebanese Muslim mountain village. In the same

village, and as long as pottery vessels continued to

16 H. Seeden, "Nustell," in Weiss (supra n. 12) 692-95; M. Krafeld-Daugherty, "Observations on Children's Burial

Customs," in Wohnen im Alten Orient (Diss. Miinster 1992).


serve as main food containers, specific traditions survived associated with a belief in the positive power emanating from the household's storage jar. Thus,

girls on first menstruation embraced the large-bellied flour jar in an attempt to ensure their future status as mother.

A recent children's cemetery was accidentally discovered in the immediate vicinity of Nustell. When

asking permission of the mukhtar's son to test a small tell along the Wadi Rijle, we learned that until the late 1970s when the inhabitants had abandoned the vil-

lage, they had used the small raised area in the wadi to bury only the stillborn and very small babies of their community. Burials took place immediately after death had occurred, and the little graves were not marked in any way. Although a special burial place, this one was not intended to be recognizable to others. We were unable to find out whether specific rites or

prayers were performed at that child cemetery. The

villagers at Nustell are Sunni Muslims; they said that this infant burial practice is common in the area, but the subject is not easily broached. Without our intention to test the site, we would never have known about its special use.

The existence of separate burial grounds for infants

points to beliefs and traditions that were obviously not invented in Iron Age Phoenicia, but appear to be as old as infant mortality itself, which was and still is much higher in areas lacking hospitals and accessible medical technology. In these situations, special rites and votive offerings exist to placate the powers of life, not only to give life, but to preserve it from the manifold dangers of pregnancy and early infancy.

Sabi Abyad. Peter M.M.G. Akkermans, National Museum of Antiquities, Leiden, reports:

The third and fourth seasons of excavation at Tell Sabi Abyad, situated about 30 km south of the Syrian- Turkish border in the fertile, undulating plain of the Balih River, focused upon the occupation levels of the sixth millennium B.C., and the late second-millennium, or Middle Assyrian, phases of settlement.

At present, 11 prehistoric phases or main levels of

occupation are recognized. The earliest levels of oc-

cupation (11-7) have been reached in a series of

narrow but deep trenches along the slope of the mound. Virgin soil was reached at a depth of over 4 m below the level of the surrounding fields. Settle- ment at Tell Sabi Abyad seems to have begun around 5800/5700 B.C. The pottery from this earliest level consists mainly of coarse vessels, occasionally with loop handles, which closely resemble the ceramics from nearby Tell Assouad and Tell Damishliyya. The pottery of the subsequent levels 10-7 also consists mainly of irregularly shaped, thick-walled, and often burnished ceramics with little variety in shape. Most

common are simple plain-rim bowls, holemouth pots, and jars with low necks, some of which showed incised or impressed patterns of crosshatching, oblique lines, or herringbones, or bands of dark red paint. Husking trays with shallow ridges or finger-impressed pits on their interior bases are found, as well as red-burnished

pottery and small quantities of imported Dark-Faced Burnished ware.

The topmost levels (6-1) have been excavated over a considerable area, up to 800 m2 or more. Earlier excavations had exposed a series of Early Halaf oc-

cupation levels, which appeared immediately below the topsoil and which can all be placed in the later sixth millennium. Interestingly, these Early Halaf strata were directly preceded by earlier Neolithic occupation levels at the site. No gap in occupation is indicated and the Halaf period at Sabi Abyad was

apparently the result of a gradual and continuous local process of cultural change. This transition seems to have taken place within a short period of time, i.e., hardly 100 years (ca. 5200-5100 B.C., uncalibrated), and is represented by three levels (6-4), the most

interesting of which is undoubtedly the level 6 Burnt

Village (the earliest of the transitional levels). The level 6 settlement consisted of dense clusters

of large rectangular houses surrounded by smaller circular structures, which all had been burned in a violent fire around 5200 B.C. (fig. 7). Thousands of

objects were recovered in situ from this Burnt Village, including ceramics, flint and obsidian tools, figurines, sealings, labrets, axes, bone implements, and jewelry.

The Burnt Village was built in terraces: along the

slope, part of the mound was dug away and the floors and walls of the houses low on the slope were founded about 2 m below those of the houses somewhat higher on the mound. It thus appears that the floors of the

upper houses were more or less at the same level as the roofs of the lower houses; one could easily walk onto the roofs. Actually, we have evidence that various kinds of activities-some apparently of ritual nature- were carried out on the roofs.

The houses of the Burnt Village were very regularly built of tauf (pisd), although all kinds of renovations and reconstructions took place over time. Some of these houses must have had 15 or more rooms, all

very small and varying in size between 3 and 5 m2.

Some of the rooms had normal doorways, while others had entrances of such restricted size that one had to crawl on hand and knees through them. Occasionally no passage at all was present; apparently these rooms were accessible from the roof only.

Ovens and fireplaces are virtually absent from the houses. These features are mainly found in the open courtyards surrounding the buildings or constructed in small auxiliary structures. One of the ovens was


0 m

~~i Y

Ijf i

Fig. 7. Sabi Abyad. Isometric reconstruction of the Burnt Village (ca. 5200 B.C.).

beehive-shaped, and stood to a man's height with the roof almost completely intact.

In the open areas surrounding the houses, circular structures up to 5 m in diameter were found. The

buildings were covered with white plaster and had small doorways. The largest of these circular buildings was divided into a series of smaller compartments, some of which had bins standing on the floor.

The floors of the houses of the Burnt Village were littered with artifacts and goods of all kinds. In one

room, several cubic meters of burnt wheat were

found, undoubtedly representing a burnt stockpile. Most of the ceramics of the Burnt Village were coarse

products, manufactured with little care in a rapid way. Bowls were typically oval in shape. A considerable

part of the pottery was incised or painted. Interest-

ingly enough, in shape and decoration many of the

painted or painted-and-incised ceramics closely resemble Samarra or, to a lesser extent, Hassuna pottery from northern and central Iraq; some kind of interaction seems to have existed between the local late sixth-millennium communities of northern Syria and those found further east. It is stressed that the Sa- marra and Hassuna traits precede the occurrence of true Halaf at our site.

In one of the houses, hundreds of finds were largely concentrated in one very small room only. This room was not used for common domestic activities and was not a storage room in the usual sense, but seems to

represent a kind of archive. Some of the most exciting finds were clay sealings with stamp-seal impressions, of which over 150 examples were found in this room,

and tokens, calculi, which point toward a very early but well-developed system of recording and administration. Another 150 sealings have been uncovered in one of the other houses. Prior to this discovery, the earliest-known sealings in clay were from the later Halaf period, i.e., the early fifth millennium, but the Sabi Abyad sealings date from the late sixth millen-

nium, ca. 5200 B.C. In general, true sealings had been

thought to originate in the Ubaid period in the north of Iraq, but this now appears to be incorrect.

In addition to the investigations of the prehistoric settlement, considerable attention has been given to the Middle Assyrian layers of occupation (13th cen-

tury B.C.) present at Tell Sabi Abyad. A monumental

building, most likely a fortress, was found on top of the mound (fig. 8). The building is oriented north-

east-southwest, is about 23 m long and 21 m wide, and has thick mudbrick walls, each 2.50-2.75 m wide and partly preserved to a height of 3 m. A staircase

gave access to a second story. Below this staircase a corridor with a vaulted roof was found.

Apart from the well of the staircase, the ground floor of the fortress consisted of eight rooms, varying in size. In some rooms, many ceramic vessels were found on the floor, especially jars and pots, as well as burnt barley undoubtedly once stored in the large jars. Most likely, these areas served as storage units. Other rooms contained a large white-plastered basin

dug out in the floor or showed a wall with a series of small niches constructed at floor level. In one of the

rooms, a carefully sawed right femur of an elephant was found. A similar elephant bone was uncovered in


Fig. 8. Sabi Abyad. Middle Assyrian fortress.

a layer of debris next to the building. All rooms were accessible through rather low and narrow doorways, consisting of carefully constructed mudbrick arches. These vaulted doorways were about 1.6 m high and about 60-90 cm wide.

The fortress seems to have had several phases of

occupation and renovation. Apparently, the building stood repeatedly empty for considerable periods of time. In the later 13th century B.C., Tell Sabi Abyad may have been part of a series of Assyrian border

garrisons, serving as a small economic-administrative center in the far periphery of the Middle Assyrian Empire.

THIRD MILLENNIUM B.C.

Abd. Uwe Finkbeiner, Universitit Tuibingen, re-

ports: Tell el-Abd is situated about 100 km to the east of

Aleppo on the eastern bank of the Syrian Euphrates lake (Buhairat al-Assad). In the fall of 1992 excava-

tions were begun by the Altorientalisches Seminar of

Ttibingen University.'7 The first excavations on the site had been carried out in 1971 and 1972 under the

directorship of A. Bounni, Department of Antiquities of the Syrian Arab Republic. In a north-south trench of 10 x 60 m, he exposed levels of Hellenistic occu-

pation, which were confined, however, to the upper parts of the mound. Below were several layers dating to the Middle Bronze Age (MBA) and to Early Bronze

(EB) IV. The 1992 campaign extended these findings con-

siderably. At the northern end of the tell (squares 18/28-19/28) a stone wall had been cleared by the waters of the lake; it was built out of large blocks almost 2.5 m wide and up to 2 m high. Pursuing the wall into the tell (squares 19/27, 19/26, and 20/26) brought a surprise: the stone wall was merely the foundation for a larger mudbrick wall that continued to rise to more than 4 m, and, with bastions and other later additions, was 10-12 m wide.

17 For earlier excavation reports, see A. Bounni, "Prelim- inary Report on the Archaeological Excavations at Tell el- Abd and Anab al-Safinah (Euphrates) 1971-1972," AASOR 44 (1979) 49-61; K. Toueir, The Syrian Archaeological Ex-

pedition to Tell Al Abd Zrejehey: Clay Figurines of the Third Millennium B.C. (Syro-Mesopotamian Studies 2.4, Malibu 1978).


It is obvious from the stratigraphy that this fortification wall goes back to the foundation of the city, which is dated by the accompanying finds to the middle of the third millennium. The wall protected the EB III upper city, including a public building of some extent, whose walls (more than 2 m wide) appear at the northwestern edge of the site.

The lower city continues to the east as a hardly visible flat terrace; it is not yet certain whether it was also walled in and whether it may also be dated to the EBA.

Two more excavation areas were opened in the middle of the site. In one (squares 20/18-20/20), we tried to expose the ground plans as documented by our Syrian colleagues in order to establish a reliable connection between their old and our new findings. The other area (squares 16/20-19/20), a trench con-

tinuing at a right angle to the west, was intended to

provide an overview of the occupational history of the

city from the latest layers on the mound to the earliest on the bank of the lake.

The eastern upper half of the sounding yielded remains of a private house dating to the Hellenistic era and to the time of the Roman Empire, as indicated

by the pottery. Besides terra sigillata, there are also

specimens of a corrugated, well-fired household ware, which already point to the beginning of the Byzantine period. Thus, the uppermost layers of the tell represent the same period that is known from the rock-cut tombs about 1 km to the east, as well as from the rich tomb of Anab as-Safina directly to the south of Tell el-Abd.

The preceding occupation, further down the slope, belongs to the Middle Bronze Age, marking a hiatus of almost 2,000 years. The remaining layers all the

way down to the bank represent continuous occupation extending over approximately 700 years, from the beginning of the second to the middle of the third millennium, down to the same Early Bronze III levels to which the city wall at the northern end of the tell

belongs. In addition to a large amount of pottery from the

periods mentioned above, there were a few painted Halaf sherds scattered close to the foundations of the

city wall. They may stem from Shams ed-Din-Tannira, a relatively nearby site. Alternatively, they may indicate a small Halaf settlement in the core of Tell el- Abd.

The small finds of the campaign consist chiefly of Bronze Age terracotta figurines, mostly anthropo-

Fig. 9. Abd. Anthropomorphic terracotta figurine.

morphic (fig. 9) or zoomorphic types. Some 250 such

pieces had been excavated earlier by Syrian archae-

ologists. In addition there were, for example, a very carefully worked steatite box, a jar-stopper with a third-millennium sealing, EB sherds with seal impressions, and several bone needles.

'Atij. Michel Fortin, Universite Laval, Quebec, re-

ports: During the spring of 1992, the Canadian Archae-

ological Mission in Syria, under the directorship of M. Fortin, resumed its fieldwork (after a lapse of three

years) at the Ninevite 5 site of Tell 'Atij, located in the

salvage area on the HIabur, to the south of Hasseke.18 This fourth campaign, which lasted 10 weeks (10

April-25 June), was subsidized by the Social Sciences and Humanities Research Council of Canada. Exca- vations were undertaken in two sectors: on the main tell and on secondary tells.

18 Preliminary reports on the three previous campaigns have been published in Syria 65 (1988) 139-71, Syria 67

(1990) 219-50, and 535-67.


Fig. 10. 'Atij. General view of level IX.

Digging on the main tell concentrated on its middle

part where five levels were identified in 1988. Four more levels, VI-IX, amounting to a deposit of 2.5 m of debris, were uncovered on a 15 x 15 m surface.

Level VI, like the upper levels, comprises a series of structures of various sizes, which, to judge from their floor and walls covered with gypsum plaster, had

obviously been designed for storage purposes; in one case, the roof of the structure was vaulted. Remains from this level substantiate the interpretation previously suggested for the main tell of 'Atij: the site was a trading outpost or a redistribution center. We cannot as yet determine the functions of the remains found in lower levels.

Levels VII and VIII consist of several superimposed, thick layers of gray ash up to a thickness of over 1 m in places. These levels would have been formed by the accumulation of ash thrown into a large courtyard with many beaten earth floors. Soil samples have been taken from a balk in order to assess

(through the use of thin sections) how these layers were deposited. Additional soil samples were taken and floated; many botanical remains were collected and will be analyzed. As in previous levels, quite a few

stratigraphical units or lots in levels VII and VIII

yielded sherds from both Metallic and Ninevite 5 wares, confirming the contemporaneity of the two fabrics, a relationship already observed on other sites

in the area. No building was found in association with the courtyard of levels VII-VIII.

Structures reappear in level IX (fig. 10); they are

adjacent to rectangular rooms opening on a courtyard located exactly underneath the one found in upper levels (VII-VIII). The beaten earth floor of this lower

courtyard has been partly uncovered from beneath a

deposit of 1.2-1.4 m of ash and debris. Room walls were made of poor-quality mudbricks originally thought to be some kind of pise.

Lastly, three more levels, X-XII, were identified in section only, during our sectioning of the remaining slope (2.4 m vertical) down to virgin soil, on the west flank of the main tell. Thus, the entire stratigraphy of the main tell has been established; the three lowest levels will be excavated in the course of our next and last campaign. We already know that level X should reveal a structure built according to a grill plan- usually associated with storage facilities-as we have uncovered the corner of such a building.

In addition to this area, we proceeded to the north of the main tell by clearing a portion of a high (5 m) and massive (1.5-m thick) mudbrick wall. Since it was laid on virgin soil, this wall would have encircled the main tell from the time of its foundation.

The secondary tell of 'Atij proved to be deceptive. Contrary to what we had imagined, houses with poor- quality mudbrick walls that were located there proved


to be of a later date than the Roman period. Some walls, unearthed this year, overlay broken Roman baked bricks that had been used for building tombs

postdating the Roman occupation of the Habur valley. Three earlier tombs belonging to the Ninevite 5 period were discovered. The tombs from this period discovered on the secondary tell now number 12. It

appears, therefore, that this area must have been used as a necropolis by the persons who operated the Nin- evite 5 trading outpost on the main tell of 'Atij.

A narrow (1 m) and long (75 m) trench was sunk between the excavated area of the secondary tell and its southern end. No trace of human occupation was noted. The entire deposit is natural, and it corre-

sponds to the tail of the island formed on the river in

antiquity and evident as a secondary tell today. Brak. David and Joan Oates, University of Cam-

bridge, report: The 13th season in the current series of campaigns

at Tell Brak began on 19 March and ended on 27

May 1992.'9 Excavation in the 1992 season was concentrated within the large Akkadian public buildings in areas FS and SS; a smaller project continued our

investigation of a well-stratified residential sequence dating to around 3000 B.C.

The Akkadian buildings. In area SS considerable additions have been made to the plan of the ceremonial complex, both completing the internal plan and extending it to the west for a further 15 m, where there is a massive tauf wall up to 5 m thick. This

certainly represents the latest identified phase of construction and was apparently a boundary or terrace wall marking the western limit of the complex, but its

upper surface at the eastern end was flush with the

secondary doorsill of the west entrance to the complex (room 21), which apparently also served as an access

ramp. From the north wall of room 21 a second, curving wall projected northward, and seems to have been an inner boundary wall enclosing an upper terrace. More work is needed here to elucidate the function of these walls and their relationship to the

previously excavated buildings. A second "kitchen," or workshop, was also excavated (room 28); it contained a very large oven similar to the one found

previously in room 4. Adjacent to it was a large stair- well (room 29), which presumably led to the roof of the complex. An unusual feature in the packing below

the stairs was the presence of some 15,000 sling bullets. Other structural relationships within the complex were investigated in a series of soundings that show that it was built piecemeal over a considerable period of time, but we see no reason to change our original ascription of it to the reign of Naram-Sin. Only the bent-axis temple, its earliest and only conventional

Mesopotamian feature, may be of slightly earlier date. In this context, it is particularly interesting that in

area FS we found this year another, smaller, bent-axis

temple in new excavations at the northeast corner of the tell. This was approached from a courtyard on the east side, now heavily eroded, which appears to be directly connected with the large courtyard excavated in 1991. Although it undoubtedly coexisted with the rest of the complex, the bent-axis shrine seems,

again, to have been the earliest feature. Its courtyard showed marks of very heavy burning, and in the small excavated area we found a number of bullae with numerical notations and seal impressions of late Early Dynastic (ED) (Sargonid?) date. These were obviously records of deliveries to the temple, and two of the

impressions bear inscriptions identifying the deity as Sakkan, who is described in a passage from the seventh (cuneiform) tablet of the series maqlu as "fertil-

izing the ewe with her lamb, the gazelle with her fawn, the donkey mare with her foal." It is thus particularly interesting that a group of gazelle horns was found on the floor of the temple cella, and that the most

conspicuous component of the ritual deposits throughout the FS complex, laid down before the deliberate filling of the buildings that we attribute, as in area SS, to a brief Akkadian reoccupation of Tell Brak, was a number of donkey skeletons.

One of the most interesting discoveries in area FS was a deposit of copper/bronze sickle blades, tools, and a mirror, on top of the obviously ritual fill of the new FS shrine. Valuable additions were also made to our now large corpus of Akkadian pottery.

Area TW. Our other principal concern was to continue the excavation of area TW, in order to elucidate further the sequence of occupation spanning the end of the fourth and the beginning of the third millennium B.C. This had produced some of our most

interesting and important results. In 1991 we excavated here some nine building

phases consisting of small houses with riemchen walls.

19 This work was generously supported by grants from the National Geographic Society, the McDonald Institute for Archaeological Research, the University of Cambridge, the British School of Archaeology in Iraq, and the British Acad- emy. For previous reports, see D. Oates and J. Oates, "Ak- kadian Buildings at Tell Brak," Iraq 51 (1989); D. Oates and

J. Oates, "Excavations at Tell Brak, 1990-1991," Iraq 53 (1991); J. Oates, "Trade and Power in the Fifth and Fourth Millennia B.C.: New Evidence from Northern Mesopota- mia," WorldArch 24 (1993); J. Oates, "Tell Brak: Uruk Pot- tery from the 1984 Season," Iraq 47 (1985).


116"s a S Ina I s 22

Steinbat II

G>

G2 St iHba HI

Stsinba 11I

Gf Gg Gh G1 CI Ha Hb Hc

Fig. 11. Chuera. "Steinbau I" and its surroundings.

These yielded a very distinctive ceramic assemblage, which immediately predates Early Ninevite 5 at Brak. This season our tentative correlation of TW phase 6 with Eanna Archaic I was confirmed. Beneath phase 6 were levels containing structures with grill founda-

tions, a type known elsewhere at this time (e.g., Tell Karrana 3), below which lay this season's most significant discoveries. The first was a level containing a number of ceramic types identical with those from the site of Jemdat Nasr, including two polychrome sherds. This is the first discovery of Jemdat Nasr

polychrome ware in northern Mesopotamia. In addition to the usual range of mass-produced vessels, this level also produced a large number of unusually tall "flowerpots," a type not otherwise attested. Also

present in this level were very large painted bowls/ craters, of a type we have not previously seen. Most

interesting of all was the large building that lay below the "Jemdat Nasr" level, from which we recovered a wide range of complete vessels of Habuba/southern Uruk type. Of significance, too, is the fact that this

building would seem to be contemporary with, or at most just later than, the area CH level 9-10 structures excavated in 1984. This indigenous northern Uruk

repertoire has a very wide distribution, from the Up- per Euphrates (e.g., Kurban H6yiik) to the northern

Jezireh, and its relative date has previously been the

subject of much disagreement. Immediately below the

large structure containing Habuba-like material there can be seen the red plaster line of a substantial niched facade, which awaits excavation next season and which should add further to the important, indeed

unique, Uruk sequence now established at Brak. Chuera. Winfried Orthmann, Universitat Saar-

brucken, reports: The 17th excavation campaign at Tell Chuera

lasted from 18 August to 15 October 1992. Excavation was continued in the main sanctuary of the third- millennium B.C. (Steinbau I and II), Palace F, and the Middle Assyrian settlement.

Steinbau I and II. The center of the third-millennium sanctuary is formed by the high platform of Steinbau I (fig. 11). In 1990 we had made a sounding in the western part of this platform in order to find out how it was built. The sounding had shown that at least two building phases were to be distinguished: under the upper stone layers, which were a substructure for the later floor, there was a fill of burnt debris, and under this the phase 2 floor. In 1992 we enlarged this sounding by removing a part of the upper stone

layers with the aid of a crane. Under the fill we found the almost intact floor of phase 2 and then under it


r?::?- ---r---- ---- ii mb

i i- -'

-

---

ftam-

Cf Cg Ch cl cl Da Db Dc

Cg Ch Ci C[ Da Db

Dc

Fig. 12. Chuera. Palace F.

again a stone substructure. Contrary to our expecta- tions, we were not able to identify stones belonging to the walls of the building that had once stood on the

platform. North of the central platform we uncovered more

rooms of building level 7. A building of phase 7b with a niched front wall on its south side (rooms 6, 102, 103, and 117) was completely uncovered. To the east of this building there is a courtyard, with a narrow

passage leading to the south. Room 118 to the west of this passage was a storeroom containing several

large pithoi fixed in their position by a filling of mud between them; this filling was plastered with white

gypsum. Room 115 to the south of this complex was

probably built in phase 7b; at that time its entrance was on the west side in a niched wall. Later in phase 7a this door was closed, the floor was raised consid-

erably and six large ovens were built on it, and a new door was opened in the north wall.

In the space between Steinbau I and Steinbau II we continued to investigate the stratigraphical and func- tional relationship between these two buildings. In the 1990 campaign we had established the fact that Steinbau II was not an Antentempel as we had assumed before, but rather a kind of gate building: it had almost symmetrical entrances on its northwestern and southeastern sides. A courtyard, paved with gravel, lay between the northwest entrance of Stein- bau II and the latest stairway leading up to the plat-

form of Steinbau I. The pavement of this courtyard corresponds to that of level 6 north of Steinbau I. The 1992 excavations have shown that this courtyard is limited on its south side by a large mudbrick wall

connecting the western corner of Steinbau II with the southeastern corner of the outer terrace on the south side of the central platform of Steinbau I (South Terrace II). The stone base of this mudbrick wall continues to the east under the southwest wall of Steinbau II; thus, it became clear that Steinbau II is a rather late addition to the sanctuary.

Palace F. In the palace area, we first removed the remains of level 1 in all those squares previously excavated. This level consists of rather small houses and rooms, apparently workshops, built together with a kiln in the ruins of the deserted palace, mainly in the former courtyards. In this level we again found much pottery, representing the characteristic shapes of the Tell Chuera IE period. From the lowest floor of level 1 came a Late Akkadian cylinder seal with a

representation of the sun god, which provides a good dating for this level.

Several new rooms were added to the plan of the level 2 palace (fig. 12). From courtyard 3, on the east side, stairs lead up to anteroom 7, from which one can reach courtyard 8 between room 6, excavated in 1985 (at that time called the "Westtempel"), and room 12, which, having a dais built against its east wall, might have been a throne room. A door in the south-


east corner of room 12 leads to a series of small rooms, one of them (21) a bathroom carefully paved with stones. The south wall of rooms 14-26 is the southern limit of the palace building, at least in its latest build-

ing phase (2a). In several rooms we were able to

distinguish two floor levels. Within the pottery found in the fill under the upper floor, we did not find the

typical shapes of the Tell Chuera IE period, but those of Tell Chuera ID; apparently the palace building is earlier than we had assumed.

In squares G.VI:f3-g3 we had started a step trench in 1990. In 1992, we continued work in these squares in order to examine the earlier building levels of the

palace. We cut through the massive, more than 6 m

thick, mudbrick wall that delimits the palace's west side. Digging down almost 4 m from its top we did not reach the base of this wall. In the fill to the east of this wall we could identify two main earlier building phases of the palace (phases 2b and 2c). The walls of both phases stand to a considerable height. The pottery sherds coming from the debris of phases 2b and 2c also belong to the Tell Chuera ID period. We went down to a deeper level, where Tell Chuera IC-period pottery could be found, only in front of the outer face of the wall. This huge wall on the west side of the

palace probably was part of an inner defensive system surrounding the citadel mound in the center of Tell Chuera.

The Middle Assyrian settlement. The intention of this

year's excavation in the Middle Assyrian settlement was mainly to complete the plan of a large mudbrick

building in level III, which was partially uncovered in the last four campaigns.

To achieve this, we had first to clear out the walls of level II, which were built on top of the ruins of level III. Level II can be subdivided into six phases. Phases 1-3 are very disturbed by modern graves. In level 11/3 and 4 we found the shafts of four Assyrian graves, which had been cut into the earlier phases 5 and 6. One of these graves was a pot-burial, while the

other three graves are mudbrick cists. In one of these

graves we found the skeleton of a woman with 100 frit beads of different sizes around her neck. On the chest lay a beautiful bone comb with a double row of

teeth, the middle being decorated with incised zigzag lines.

Within the large building of level III three phases can be distinguished. Phase 2 is the most important of them. The main entrance is situated in the south-

east of area G. It is flanked by two towers; two door sockets were found in the entrance room Ic. From here one can enter the very large room 1. Through a small doorway opposite the main entrance, one reaches antechamber 7 of the large (central?) courtyard 8, which had a stone pavement. From room 7 two doors lead to room 2, the room in front of archive room 3. In front of a blocked door to room 10, in which we only found one big vessel in the southern corner, and in a kind of niche in the northeast wall, we uncovered nearly 40 tablets and a number of fragments and envelopes of the Middle Assyrian period, some of them containing limu-datings from the time of Tukulti-Ninurta I (1244-1208). Among the tablets are letters, which apparently had been sent to the local Assyrian governor residing in this building. The texts mention the city of "Harbe" (or "Hurbe"), and this place-name, which also occurs in a text from Dur Katlimmu, probably can be identified with Tell Chuera.

Gudeda. Michel Fortin, Universit6 Laval, Quebec, reports:

A third campaign, undertaken by the Canadian

Archaeological Expedition to Syria and directed by M. Fortin, was held in the spring of 1992 (10 April- 25 June) at Tell Gudeda, a small site located in the salvage area on the Middle Habur to the south of Hasseke, dating from the end of the Ninevite 5 period or perhaps even the beginning of the Akkadian period.20 This campaign was subsidized by the Social Sciences and Humanities Research Council of Can- ada. Excavations were carried out on top of the tell and at the bottom of its southern flank.

The main area of excavation was on the flat summit of the tell where during previous seasons several 5 x 5 m squares had been opened. This year operations continued in seven squares. The amount of earth removed varied from several centimeters to a depth of 2 m, according to the importance of preserved remains. Two new levels were identified (II and III).

In level II, the more recent of these levels, we have distinguished three distinct building phases; structures from this level were apparently restored and reused very frequently. A few rooms, however, remained in use through all three phases, such as room 500 (fig. 13), the four massive walls of which are covered with gypsum plaster bordering a trapezoidal area of 3.5 x 3.8-4.5 m. Inside room 500 were various installations: a thick, arched buttress and four smaller

20 Preliminary reports on the three previous campaigns have been published in Syria 65 (1988) 139-71, Syria 67

(1990) 219-50 and 535-67.


Fig. 13. Gudeda. Room 500.

ones, low benches at the bottom of walls, and three basins set right into the plastered floor. The latter was made of several layers of plaster indicating a prolonged occupation of the room. Although we do not know the purpose of this room, it is worth mentioning the discovery, on the floor, of about 4,500 shells stored in ajar (of which we retrieved fragments) before being perforated with a metal pin found nearby. The rest of level II consists mainly of a large courtyard, about 10 x 10 m, surfaced with pebbles; three phases of these pebbles were superimposed. This courtyard also included plaster ovens and basins located next to small rooms. These installations give the impression that this courtyard was not reserved for domestic use.

Furthermore, excavations in this area revealed a ceramic waster and a fragment of what seems to have been a terracotta crucible still bearing traces of molten metal. Lastly, the presence in this level of two bone

fragments, each regularly notched on one narrow

side, deserves mention. Two building phases were identified in the under-

lying level III. The actual construction of room 500, described above, is dated to this level. The large court-

yard (also mentioned above) was in use at this time

although its floor was simply made of beaten earth rather than pebbles as in the upper level; it also included some tannurs, or bread ovens.

At the bottom of the southern flank of the tell, which gently slopes toward surrounding fields, we

opened a 5 x 15 m trench to determine the limits of settlement in that direction. Contrary to our expec- tations, the limits of the site were not reached nor was

virgin soil. We conclude that the occupation of Tell Gudeda must have extended further in the direction of fields where modern agricultural operations have erased all traces of the site. Architectural remains were unearthed, but as they were very near the sur-

face, they have been greatly damaged by roots. None-

theless, a few structures have survived: a floor made of mortar with a bitumen patch and an oven that

belonged to a 3.5 x 4 m room, enclosed on three sides by mudbrick walls with bases of stones. The lower portion of a stone staircase remained in a corner of this room. This area yielded Ninevite 5 painted ceramics, fragments of pottery decorated with incised animal outlines, and a bronze chisel. Another poorly preserved room with earth walls had been erected to the north at a higher level.


Several layers are visible in the balks; they will be related to the stratigraphy observed in the trench

opened in 1988 along the northern slope of the tell. We also hope to be able to link the trench opened on the southern flank with the area excavated on top of the tell in order to obtain a north-south section across the whole of Tell Gudeda.

Khirbet el-Umbashi. Frank Braemer, Institut

frangais d'archeologie du Proche-Orient (IFAPO), J.-C. Echallier, CNRS, and Ahmed Taraqji, Director- ate General of Antiquities, Damascus, report:

Under the joint direction of the Directorate General of Antiquities of Syria, CNRS, and IFAPO, excavation was begun in 1991 at Khirbet el-Umbashi, within the framework of the research program entitled "The Peopling and Development of Hauran (Southern Syria) from the Chalcolithic Period to the Coming of Islam."

The site is located on a volcanic hilltop about 30 km east of the last sedentary villages of the Jebel el- Arab region, and 80 km southeast of Damascus. The region is now a desert.

The main period of occupation dates from the second half of the third and the beginning of the second millennium (EBA III, IV, and MBA I). Epi- sodes of Chalcolithic and EB I/II settlement, however, are also present.

The built-up area, which covers approximately 60 ha, has an uncommonly rich animal bone deposit with the remains of goats, sheep, bovids, and gazelles spread over a surface area of approximately 3,600 m2. The site does, however, include numerous other remains: several hundred dwellings divided into three distinct and adjacent groupings, which for conve- nience we have called North Town, South Town, and West Town, and two necropolises to the east and southwest comprising over 1,000 built-up tombs, mostly of dolmen type. On the temporary river that borders the site to the north, there are vestiges of a dam and of a great reservoir that had trapped a layer of sediment several meters thick, carried over by flooding from the wadi. A system of canals is also visible.

No earth has accumulated over these traces, allowing us to follow their contours quite easily and to understand their architecture. All the buildings are in an exceptional state of preservation, allowing a close analysis of the building systems, which were based on the post-and-lintel principle.

Two fieldwork expeditions were undertaken in 1991 and 1992. Their aims were 1) to produce a full topographical plan of the constructed area as well as

detailed architectural surveys in order to carry out an urban and architectural study; 2) to excavate the site in order to date the various adjacent elements and establish an internal chronology for them; 3) to collect information on the environment of the time in order to understand the reasons for and possibilities of a settled community; and finally 4) to chart all the traces of human settlement on a microregional scale in order to describe the settlement patterns better.

The necropolis. More than 1,000 dolmen-type tombs were built under a tumulus and enclosed within a

quadrangular low wall. This necropolis has no point of contact with the dwellings, which makes it impossible to give the relative dates of the two areas. More- over, its size leads us to believe that it could have had its own evolution independent of that of the occupied zone. For example, it could have been a central ne-

cropolis for the non-sedentary peoples of the region, whether or not they were contemporaneous with the

phases of settled communities. Four adjacent and architecturally successive tombs

have been excavated. All include an exterior low wall

containing the upper part of the fill of pure earth surmounted by stones, forming the tumulus of the tomb. The tumulus covers the cave itself, closed by means of one or two great slabs supported either by flat stones set on edge, or by a low wall. At the level of the cave opening, we found the remains of food

offerings: bird bones mixed with charcoal and earth. The bodies were left at the bottom of the cave. There

may have been a period following the time of the

deposit when the cave stayed empty. It was finally filled with alluvial mud, consolidated and tamped, no doubt, with water.

South Town. The South Town covers a surface area of approximately 7-8 ha. It was built on bare rock left by lava flow and consists of several distinct elements.

The ossiferous deposit mentioned above varies in thickness from a few centimeters to more than 2.5 m. The bones have been butchered and burnt. In some

places, there are blocks of melted basalt incorporating pieces of bones. These two phenomena, a huge accumulation and incineration, have not yet been ex-

plained; however, they are certainly human in origin. The excavation established that these accumulations are certainly secondary deposits such as dumps of heterogeneous refuse.

The study of the fauna displayed the presence of 75% ovicaprids, 24% bovids, and 1% gazelles. Al- though sheep and goats could be expected in large numbers, the proportion of cattle is surprisingly high


and abnormal in a desert area, as cattle need grassland and water in great amounts.

A rampart wall can be traced for almost 500 m to the north, east, and south of the town. It is not ho-

mogeneous either in its dimensions or in its construction. We can distinguish at least four phases, all of them structurally associated with a system of dams on the wadi, a protected water reservoir being thus provided. The area enclosed within the rampart was

basically empty except for a 25-m long fortified house. In the western part of the South Town, the sur-

rounding wall was apparently destroyed and replaced by a very unusual 160 x 120 m area of grouped dwellings. These houses are partly underground, with rounded contours and connecting rooms, suggesting an agglutinative type of architecture and town plan- ning.

Twenty-three houses containing about 60 rooms have been located, measured, registered, and drawn. Each house consists of two inhabited floors, one above the other, with almost independent entrances and circulation. The upper floor is, for the most part, in the open air and serves as a courtyard for the unit. The lower floor is underground and practically devoid of lighting, with sunlight coming in only through small doors. In about 60 rooms, the roofs have remained, formed of great slabs of basalt supported by posts and corbeling. The number of rooms in a house varies from two to 10.

North Town. A second constructed area has been discovered immediately north of the wadi. This

grouping is far less dense than that of the South Town, but covers a minimal surface area of 50 ha. The dwellings consist of groups of more or less com-

plex houses. The general arrangement follows a

roughly north-south orientation with buildings laid out in an orthogonal pattern. The arrangement as a whole does not really display an urban character, but rather that of a very large village.

The plans of the houses do not vary greatly. They are rectangular houses with axial pillars, so-called "broad houses," with a door on one of the longer sides. There are more than 400 basic cells. Most of the units consist of one room 2.5-3 m wide and 8-10 m long. The ages of the buildings themselves have not yet been determined. There were few finds from the excavation, mainly a small number of unidentifi- able potsherds.

The Wadi Umbashi and hydraulic work. The site depends entirely on the winter floods from the Wadi Umbashi (which flows from southwest to northeast) for its water supply.

There are numerous installations on the wadi in addition to the ramparts described above in the South Town. Immediately above the dams, rows of blocks disposed at the bottom of the river were meant to deviate the power of the stream in order to protect the dams. A canal over 1.5 km long was added on the left bank, which collected the rainwater running down the northern slopes and the water that dripped from a small adjacent reservoir.

Two kilometers above the South Town, remains of another stone dam were identified. The purpose of this dam was to raise the water level to feed an off- take on the right bank that flows into a constructed canal. This canal, approximately 200 m long, opens onto a succession of connected hollows in the basalt flow and ends in a large closed hollow that forms the main reservoir, situated only 500-600 m from the South Town. This hollow is 20-100 m wide and 500- 600 m long.

Environmental conditions. A pedological study of the area was carried out in an attempt to determine the characteristics of the environment at the time of oc-

cupation, and to see whether modifications have occurred since. Sections through the wadi were made, as well as sections through the early limons that cover most of the area, and the great hollows situated above the site. Two soundings were also made into kite walls; these structures are usually dated to the Neolithic period. The collected data lead us to believe that the natural environment in the third millennium was not very different from what it is now. The fact that the region is most unlikely to have been regularly culti- vated raises the question of food provisioning for the inhabitants.

Conclusions. Excavation has thus shown evidence of several superimposed or adjacent phases of occupation across the site, whether dwellings or necropolises. The detailed study of the houses of the southern and northern parts of the site shows that the types of occupation vary greatly: the large amount of ceramics, milling tools, and ornaments, and the agglutinative architecture in the south are very different from the northern built-up area, which is characterized by a lack of finds, and has rectangular houses set orthog- onally.

Finally, the discovery of a new site near Khirbet ed- Daba proves that the phenomenon of sedentary occupation in the semidesert steppe of this region is not an isolated event related to the EBA/MBA transition.

Leilan. Harvey Weiss, Yale University, reports: From 1 September through 15 October 1991 the

Yale University Tell Leilan Project concluded the first-


stage fieldwork initiated in 1978 and tested ap- proaches for second-stage fieldwork to commence in

1993.21 Toward the completion of first-stage definition of Leilan settlement history, four excavation operations were undertaken:

1) The southwestern "lobe," tested at Operation 6, was suspected to be the period I karum at Leilan, documented epigraphically within the Leilan Lower Town palace archives, but proven to be part of the Leilan IIId-IIb city founded upon calcic virgin soil and dry river/wadi beds. Following the abandonment of the third-millennium city, this area was vacant until Leilan period 0, now defined as post-Leilan I and close in time to Nuzi ware assemblages elsewhere.

During this period the mounded surface tested at

Operation 6 was a cemetery: high-status inhumations with ceramic, copper/bronze, and exotic stone burial

goods were retrieved within the 250-m2 test area.

2) Operation 7, a 25-m2 test of the northeastern

quadrant of the Leilan Lower Town, provided strati-

graphic, ceramic, and architectural evidence for oc-

cupation from Leilan IIId, founded upon calcic virgin soil, through Leilan IIb, followed by a period I administrative building set upon the leveled collapse of

period IIb architecture. Period 0 elite burials were sunk into the ruins of the period I building. The

expansion of Operation 7 to 325 m2 exposed 12 in- terconnected rooms of the period I building. On the floor of room 12, four period I vessels, broken in

antiquity, contained 590 limu-dated administrative

texts, mostly the receipts and disbursements of a beer

official; the disbursements bear the seal of a servant of Qarni-Lim of Andarig.

3) Operation 8, a 25-m2 test of the northwestern

quadrant of the Leilan Lower Town, produced an initial Leilan IIId occupation set within trenches cut into calcic virgin soil, followed by a sequence of Leilan

II occupational strata of ash and refuse, an occupational hiatus, period I exterior surfaces and a pit extending down to virgin soil, and a large period 0 ceramic kiln filled with wasters, slag, and other kiln debris.

4) An Operation 4 extension, intended to locate,

define, and date the third-millennium city wall, was

excavated mechanically and by hand to virgin soil

through period I and period II stratified deposits. A

defensive system of three contemporary parallel walls, built within trenches cut into calcic-horizon virgin soil, was uncovered. The wall construction trenches and

floors are tied stratigraphically and ceramically to the Leilan IIb period.

Toward the initiation of second-stage Leilan field-

work, which will be directed at period-specific archae-

ological problems, a research effort with M.-A. Courty (CNRS, Paris) was undertaken to test the accessibility of landscape and climatic data for Leilan and the

surrounding region from the sixth through second millennia B.C. This program entails soil micromor-

phology, on the microenvironmental and regional levels, for primary or secondary evidence of landscape and climatic conditions. Among the issues attracting

specific attention in this program were landscape and climate during "Habur hiatus 1," ca. 2200-1900 B.C., between the end of Leilan IIb and the beginning of Leilan I; Leilan-region hydrography and Leilan II-I watercourse management; and Halaf to Habur period climate change.

Execution of this test program involved exposure of 1-m wide, 6-m deep, backhoe sections against and

across archaeological deposits to expose datable virgin soils featuring climate-sensitive calcitic features and

the datable stratigraphic record of landscape and

water management, soil formations, and soil formation processes. Sites situated along the Wadi Siblah were selected for testing: the Halaf-period site of Tell

Bager located 500 m north of Do Gir, the Uruk-period site of Tell Nasran located 7 km north of Leilan, and the Ubaid site of Nasran II located south of Tell Nasran. In addition, backhoe tests of paleosols, sedi-

ments, and cultural stratigraphy were executed at Leilan in four new locations, trenches B-E. These

provided new data for Leilan settlement history, hydrography, and short-term climatic change. Trench B also provided a section through the northern city

21 Fieldwork and subsequent analyses were supported by the National Endowment for the Humanities, the Dula Foundation, Leon Levy, and Raymond and Beverly Sackler. Staff included P. Akkermans, associate director; M.-A. Courty, soil micromorphologist; W. Wetterstrom, palaeobot- anist; R. MacNeill, surveyor; J. Clark, H. Fokkens, C. Ott, G. Pulhan, L. Senior, and D. Stein, archaeologists; F. Ismail and M. van de Mieroop, epigraphers. For the 1989 season of excavations and details of 1991 research, see P. Akker- mans et al., "An Administrative Building of the King of Andarig at Shubat Enlil," NABU 1991:4, 68-70; L. Senior and H. Weiss, "Tell Leilan 'Sila Bowls' and the Akkadian Reorganization of Subarian Agricultural Production," Ori- ent-Express 1992:2, 16-24; Weiss, "Tell Leilan 1989: New

Data for Mid-Third Millennium Urbanization and State For- mation," MDOG 122 (1990) 193-218; Weiss, "Chroniques des fouilles: Tell Leilan," Orient-Express 1991:2, 3-5; Weiss, "Habur Triangles: Third Millennium Urban Settlement in Subir," NABU 1992:4, 91-94; Weiss, "Subir v. Sumer: For- mazione secondaria e collasso dello stato nelle pianure del Khabur," in Rouault and Masetti-Rouault (supra n. 3) 40- 51; Weiss and M.-A. Courty, "The Genesis and Collapse of the Akkadian Empire: The Accidental Refraction of Histor- ical Law," in M. Liverani ed., Akkad: The First Empire (Padua 1993) 129-54; Weiss et al., "The Genesis and Collapse of Third Millennium North Mesopotamian Civilization," Sci- ence 261 (1993) 995-1004.


Table 1. Subir, Third-Millennium Developmental History: Leilan Chronology, Direction of Regional Influences, and Climatic Conditions

'4C dates Leilan Habur Plains' Direction of Climate B.C. Period Developments Influence Dry I Wet

ca. 1600 0 ? ? 1900-1725 I resettlement of Subartu south to north 2200-1900 Habur hiatus 1 Subir 4: desertification/desertion north to south I I 2300-2200 IIb Subir 3: Akkadian imperialization south to north I 2400-2300 IIa Subir 2: state consolidation north to south 2600-2400 IIId Subir 1: secondary state formation south to north I 2900-2600 IIIb-c isolation/insulation 3000-2900 IIIa collapse north to south 3300-3000 IV Late Uruk expansion south to north I

walls, now identified as earthern ramparts set south of and parallel to an east-west natural rise. Datable

virgin soils and sediments were also sampled at Op- erations 1, 4 extension, 5, 6, 7, and 8.

Recent Leilan research conclusions provide for a

four-stage developmental history of the HIabur plains, ancient Subir (table 1). Particularly significant are 1) the evidence for Leilan IIId settlement provided in trench E, trench D3, Operations 5 (Lower Town South), 6, 7, and 8, suggesting a city of ca. 100 ha; 2) the triangle of Leilan IIId-IIb urban settlement, represented by Leilan-Mozan-Brak, that controlled most of the Habur plains from ca. 2600 to 2200 B.C., with the Jagjagh serving as the frontier between Leilan and Mozan; 3) the Akkadian imperialization of Leilan IIb agro-production identifiable within Lower Town South subsistence data, "sila-bowl" manufacture, city wall construction, and river channelization; 4) the

collapse of Akkadian imperialization due to an abrupt climatic change that probably generated the syn- chronous collapse of settlement and political systems from the Aegean to the Indus; the "kingdom of Urk- ish and Nawar" was probably the remnant, post-collapse, Subarian state; and 5) the period I city, Shubat Enlil, considerably smaller than the period IIId-IIb city, did not include the Leilan southwestern "lobe."

Mardikh/Ebla. Paolo Matthiae, Universita di Roma, reports:

The 28th and 29th excavation campaigns at Tell Mardikh-Ebla took place respectively between 10 Au-

gust and 30 October 1991, and between 8 August and 1 November 1992.22 In 1991 the operations were

carried out in three sectors: area P West in the Lower Town, yielding results concerning the large MB II cult terrace called monument P3; area P South in the same region of the Lower Town, with a series of EB IVA rooms; and area G West in the western region of the acropolis, yielding data concerning mainly the EB III, MB II, Iron Age II-III, and Persian periods. In 1992, excavations were undertaken in areas P West, P South, and G West, where work had been done in 1991, and soundings were also made in four new sectors: area T South; the central and northern areas of the Lower Town Northwest; regions far from the foot of the acropolis, but close to the line of the western MB I earthenwork ramparts; and area P North, also in the Lower Town, immediately to the north of the northern limit of the MB II Northern Palace (Palace P). For the first time, important data from the Lower Town of Tell Mardikh were retrieved from area T, dating for the most part to EB IVB; in area P North, we uncovered the southern limit of a new large monumental building (building P5), dating from MB I.

For the earliest periods of settlement at Ebla, EB III-IVB, the most important results were obtained on the acropolis in area G West, and in the Lower Town in area P East (near the northwest foot of the acropolis), and in area T, which is not far from the later fortifications. In area G West, the interruption of the levels of the Royal Palace G allowed us to identify three successive levels of EB III, the first of which appears to be quite archaic. Two EB IVA reconstructions of the floors of Palace G have been

22 For previous reports, see P. Matthiae, "Tell Mardikh- Ebla (Siria), campagna di scavi 1991," Orient-Express 1991, 1-3; Matthiae, "Tell Mardikh-Ebla, campagne de fouilles de 1992," AAS (in press). See also the final reports by S. Maz- zoni, Le impronte su giara eblaite e siriane nel Bronzo antico (Materiali e studi archeologici di Ebla I, Rome 1992) 1-266, pls. I-LI; F. Pinnock, Le perle del Palazzo Reale G (Materiali

e studi archeologici di Ebla II, Rome 1993) 1-178, pls. I- LXVIII; Matthiae, "Mardikh/Ebla," in Weiss (supra n. 12) 707-709; Matthiae, "Les dernibres decouvertes d'Ebla en 1983-1986," CRAI 1990, 405-10, 417-23, figs. 12, 17-19; Matthiae, Ebla, un impero ritrovato: Dai primi scavi alle ultime scoperte (Turin 1989) 171-75, fig. 40, pls. 93-98.


Fig. 14. Mardikh/Ebla. Area P West, monument P3.

found, occurring before the final phase when the palace was destroyed by Sargon of Akkad. In this same sector, excavation revealed a massive terrace wall, at least 4.5 m high and running west-east, probably of MB I date, which must be the origin of the relatively strong slope of the inner central sector of the acropolis. The discovery of this high terrace wall now explains, at least in part, the apparent loss, to the south, of the EB IVA Royal Palace G: it seems that the palace is not really lost, but that it is at a much lower level southeast of the western edge of the acropolis, which was much higher than the inner regions during EB IVA, as well as in MB I-II.

In the Lower Town, at the foot of the acropolis, the continued exploration of the EB IVA rooms allowed us to bring to light sectors of nine rooms belonging to one architectural unit, which has been called building P4. This building, which is yielding a large amount of ceramic material in a good state of preservation, corresponds completely to the Royal Palace G and was certainly destroyed by the fires set by Sargon, but its function has not yet been ascertained. Its location in the middle of the large sacred area of MB II suggests that it may have had a religious function.

For the first time, in the Lower Town Northwest, three successive levels of the post-palatial EB IVB period have been discovered. These were identified in the soundings opened in the three sectors of area T, which were quite distant from each other. There are important remains of floors, with considerable pottery. The oldest of these floors was apparently set directly on bedrock, which is particularly high in those

regions of the Lower Town. Two aspects of these discoveries are quite enigmatic: first, the extreme paucity of EB IVA ceramics just above bedrock would lead one to believe that in these regions of the tell, the settlement contemporary with Palace G was quite modest; second, the almost complete absence of walls in the following EB IVB period indicates that the function of these sectors of the urban area was quite peculiar.

A 9 x 4 m trench was opened in area P North immediately below a large MB II pit, which might have been used during the last phase of the Northern Palace. In this trench, we uncovered the huge stone foundation of a large wall, 3.2 m thick, exactly parallel to the northern perimeter wall of Palace P. The short extension of the sounding allowed us to bring to light only a second wall, perpendicular to the first one, 1.5 m thick, and also parts of two large rooms, which led us to presume that building P5 stretches for a large area in the central region of the Lower Town North. The attention to technical details (in particular, a thick

plastered floor), and the size of these structures lead us to believe that building P5 was an important large palatial structure, most likely built during MB I, or in the first years of MB II at the latest.

Important results were obtained with the complete excavation of the imposing and peculiar monument P3, dating to MB II (fig. 14). This building, completely made of stone, with the main facade on the east side, is a large cult platform, 42 m wide and 52.5 m long, inside of which was a court 23.5 x 12 m in area, located 25 m from the east facade and 15 m from the


'I ......................

"!ill, .... Ill.I!.IN immmml'

"iF'UUii EU E

U.,i~m Fig. 15. Mardikh/Ebla. Area P South, votive objects.

west facade; the court is located in the center, along a north-south axis, with its north and south limits 9.5 m from the perimeter side walls. This building exhib- its beautiful masonry: all of the inner and outer facades have regular courses of roughly square stones of a consistent height, which had been placed pro- gressively stepping back from bottom to top. In this way, the facades of the building and the perimeter walls of the court have a slight but evident and regular inclination from bottom to top, like Egyptian mastabas or the bases of Mesopotamian ziggurats, with a value

corresponding to one-tenth of the height; thus, on a preserved height of nearly 1.75 m, the top of the walls is stepped back from the line of the bottom course by nearly 0.18 m. These data indicate that the unit of measure employed was probably the cubit of 0.52 m. This being the case, monument P3 was planned to be 100 cubits long, 80 wide, and, probably, 10 cubits high, with an inclination of the walls from bottom to top of nearly 1 cubit. If these hypotheses are correct, the architects of monument P3 were familiar with the Egyptian measures and the decimal system. The ratios they imagined between inclination, height, and length would have been 1:10:100.

The intentional filling of the court, with no trace of destruction inside and around the building; the pres-

ence, against the south wall, of limestone rubble form-

ing an inclined plane; and the absence of entrances or staircases to the platform, as well as to the court, lead us to infer that monument P3, certainly planned during MB II, was probably never completed, and was filled up (perhaps in order to prevent desecration) not long before the final destruction of the town around 1600 B.C. The function of this extraordinary MB II building is difficult to determine, because its core, namely the inner court, was probably never used. It seems likely, however, that the court was

planned to house animals or trees sacred to the deity. Functionally, monument P3 is certainly complementary to the nearby temple P2 and both were probably dedicated to a great goddess similar to Ishtar, whose

front-facing, naked, winged image, quite fragmentary, was found on a fragment of a carved basalt basin in temple P2. The attribution of temple P2 and monument P3 to the Syrian Ishtar is confirmed by the excavation of a favissa located nearby in the middle of the large square on which temple P2 stood to the north and monument P3 to the west. In it, together with a very large number of votive bowls and small

jars, several fragments of precious materials and some

complete dedicatory objects of bronze were collected, among which snakes are predominant (fig. 15). The


Fig. 16. Mari. Walls of the entrance to the third-millennium palace.

archaeological evidence seems to prove that the large area of MB II date uncovered in area P South and West was the main religious center of high Old Syrian Ebla, and that in it a goddess was worshipped who, many centuries later, in the Middle Assyrian period at Ashur, was simply called Eblaitu, "[the Goddess] of Ebla."

Mari.Jean-Claude Margueron, Universite de Paris, reports:

During the 29th season of excavation at Tell Hariri, five areas were opened, the stratigraphic sequence was investigated, and methods of mudbrick wall conservation were examined.

Palace of the Shakkanakku. Investigations in this area were undertaken in the northern section, as well as on the outskirts of the building on its western flank. Some of the archaeological levels have eroded away on the northern edge of the tell, and thus it was

impossible to find the door of the palace. By exam-

ining the remains of the foundations of the palace, it has been possible to outline the limits of a large front court associated with the missing entranceway. The exact location of this door is unknown.

The organization of the palace is quite clear, and we now have an example of the innovative architecture of the Shakkanakku (i.e., Ur III) period: the main part is organized around a vast courtyard; this

gives access to the throne room, which is the bridge between this part of the building and the private

quarters and administrative and economic sections. The evolution in architectural concepts apparent in this building, so different from the huge accomplishments of the ED period, should be further investi-

gated at Mari, and Mesopotamia in general. The study of the environment of the palace (under-

taken on the eastern side in 1987, and on the southern side in 1990) was conducted on the western side dur-

ing this season. Along a western enclosure wall, oc-

cupation levels were found together with disturbed floors, mudbrick and baked brick walls, and ovens

leaning against the wall of the palace. These are un-

doubtedly poor-quality dwellings, built after the destruction of the palace by Hammurabi in 1760 B.C. This level confirms the existence of Hana-period oc-

cupation at Mari. Entrance of the third-millennium palace. The access

system of the third-millennium palace was uncovered

(fig. 16). Its great size gives Maria unique position among the architectural accomplishments of Meso-

potamia. The architectural principles behind its construction are known from early third-millennium urban enclosure walls (e.g., Jawa, Troy II citadel, Kish), but the Mari case is unique: excavated to a

depth of 19.5 m and a width decreasing from 5 m at the entrance to 3.3 m at the third tenaille, it is of a scale exceptional for the building itself. It is now

appropriate to extend excavations to the west and east to define the arrangement at the entrance, to deter-


mine whether the system is a replica of that used at Kish, and to discover the nature and organization of the units that border the street.

The western domestic quarter (chantier F). The trench intended to link the ED quarter and the second- millennium palace was deepened at various points, and extended to the east. It revealed new information on the succession of the levels extending to the south of the street that had defined the axis of the excavation.

Three major levels have been uncovered: 1) the Shakkanakku period, with subperiods defined by architectural constructions, and intramural burials with bronze weapons, jewelry, and pottery for grave goods. A potters' workshop, with work areas, turning elements, and parts of kilns, was also found; 2) Post-

Early Dynastic period, or Early Akkadian, with a sew-

age system (similar to systems found in the area of the Ishtar temple and chantier B); and 3) Early Dy- nastic, with a characteristic domestic altar similar to those of chantier B and the Maison rouge.

The temples (chantier G). The work done in 1990 in the sacred areas was continued to complete the study of the Ninni-zaza temple. The northern foundation

deposit was found: a nail fixed vertically in a ring, accompanied by small uninscribed lapis lazuli, alabaster, and silver tablets.

Under the threshold of the door a large rectangular slab of rock was placed within the foundations. Cup- shaped and geometric designs appear on its upper face.

A deep excavation undertaken along the western flank of the temple, joining the Maison rouge, has shown heavy erosion after the destruction of the tem-

ple, and before the massive works of the Shakkanakku

period. Excavation in chantier B. In order to define better

the large building from which late ED (possibly Early Akkadian) tablets were retrieved in 1980, the excavation was extended to the west. This has permitted a clearer understanding of the stratigraphy on the summit of the tell (Shakkanakku period) immediately above the building. New rooms with complex bitumen installations have been uncovered here.

Stratigraphy. Since the resumption of excavation at

Mari in 1979, investigations have been undertaken to determine the nature of the formation of the archae-

ological tells. It seemed worthwhile to define the ceramic deposit of a building falling apart before our very eyes. Room 77 of the Royal Palace, excavated by Andre Parrot, was chosen. Its cleared levels are cur-

rently in the process of being reburied. The quantity of sherds, stones, and other stratified material origi-

Fig. 17. Mari. A "protective cap" on one of the pillars of the Salle aux Piliers.

nates essentially from the decomposition of the walls; significant anomalies present in the ceramic deposit can be identified. A reevaluation of the significance of the stratigraphy and its role in ceramic studies is the primary consequence of this type of inquiry at Mari.

Conservation of mudbrick architecture. The work undertaken since 1988 with the CRAterre (Centre de recherche sur l'architecture de terre) of Grenoble has been continued on a large scale. Numerous measurements of wall humidity and erosion resistance have been taken. Many attempts to restore the walls and the arrangement of floors and subfloors have been undertaken in the Enceinte sacrde of the third-millennium palace (level P-2), in the Dagan Temple annex north of the Haute Terrasse, in the Palace of the Shakkanakkus, in two pillars of the Salle aux Piliers of level P-1 in the third-millennium palace (fig. 17), and in the foundations of the Ninni-zaza temple.

Mozan. Giorgio Buccellati, University of California, Los Angeles, and Marilyn Kelly-Buccellati, California State University, report:

The seventh season of excavations at Tell Mozan took place in June and July 1992, under the direction of G. Buccellati and M. Kelly-Buccellati. Excavations are funded through a grant from the National En- dowment for the Humanities, with matching grants from the Ambassador International Cultural Foun- dation and various donors. Excavations concentrated on two areas where work had been started during the last season (1990), and on which we reported previously: in F1, on the northwest slopes of the High Mound, were the remains of a medium-sized structure, from which came two cuneiform tablets (now published in Mozan 2); and in AS, a stepped trench


Fig. 18. Mozan. Public building in area AS, from above.

was dug on the western side of the High Mound, where a sequence was established from the Nuzi period down to the middle of the third millennium.23

The stratigraphic sequence of the building in Fl

(dated to the Late Akkadian period because of the tablets found there) appeared relatively simple in the 1990 season, but turned out to be more complicated. The large storage room had at least two phases. In the earlier of the two, the storage room could possibly have been connected with the upper area where the two cuneiform tablets had been excavated in 1990. Instead, this upper level is connected with a series of small rooms found to the south. No additional tablets were found in this area.

As indicated in our last report in AJA (95 [1991] 712-14), the large stone building at the bottom of the

stepped trench AS gave evidence of being one of the

largest stone-based buildings known from third-millennium Syria. Our excavations this season indicated that the size is even larger than previously anticipated: what we assumed to be a perimeter wall on the north side appears now to be a large interior wall, with a number of doorways opening from the south to the north. Figure 18 shows this complex, photographed toward the end of excavations. The floor plan of the

building is quite symmetrical in its articulation, and it seems to indicate that on the east side there was a second wing with a layout arranged as a mirror-image to the part already excavated on the west; the considerable width of the wall to the north suggests that there was a major division between the southern two

wings and another wing to the north. The excavated

wing in the southwest measures about 15 x 20 m, and the complete building may be three times as large as the portion excavated so far.

While the building is impressive because of its structural layout and construction techniques (well-laid stone foundations and substructures, well-laid floors, and some extremely thick and cement-like pave- ments), it is significant that its walls were not plastered. These characteristics, together with the regularity of the layout and the nature of the finds (see below),

suggest that it was a large storehouse, most likely public in nature and possibly serving the gate, which is likely to have been located in the wall that encircled the High Mound some 30 m to the west. As for the

layout, we assume the main entrance to be on the south side (only wall foundations are left there, so we have no direct evidence; future excavations may tell if an analogous doorway is still to be found in the

23 L. Milano, Mozan 2: The Epigraphic Finds of the Sixth Season (Syro-Mesopotamian Studies, Malibu 1991).


Fig. 19. Mozan. Akkadian seal with lyre player as a secondary motif.

eastern wing); if so, the rooms that have been com-

pletely excavated to the southwest may be understood as an accession suite, in which goods were received and registered by the scribes. Just beyond this suite,

immediately south of the large transverse wall, was a

large room (possibly a courtyard) with a closet in the corner. Because the interior area is so small (a little over 1 m2) and the walls are, in contrast, quite thick, this closet (as well as its counterpart in the eastern

wing) may properly be understood as a vault. The

large number of seal impressions found in the floor

deposit in front suggests that this vault may corre-

spond to what is called in the texts E2.KISIB, "the house/room of the seal(s)," i.e., an area where sealed

containers, such as boxes or jars, were kept in safe

storage behind a sealed door. The sealings from this area consist of over 80 un-

baked clay fragments with seal impressions on them. The majority of the seals had been rolled on box

sealings but some were jar and door sealings. For the most part, the reconstructed seal designs fit into the standard Akkadian corpus. One secondary motif

placed under an Akkadian inscription (fig. 19) shows two facing, seated figures, one of whom is playing a

lyre. Another impression shows an adult figure hold-

ing a child on its lap. A total of 27 sealings are inscribed. In addition,

there is one inscribed bulla, two fragments of tablets, and one complete tablet. The latter is a school tablet, with an excerpt from the ED LU E list of professions known from Abu Salabikh and Gasur.

Qara Quzaq. G. del Olmo Lete, UAB, reports: The University of Barcelona's Archaeological Mis-

sion in Syria resumed work in 1991 by digging six

contiguous 10 x 10 m units, some of which had

already been excavated in the last season.24 They were all situated in the upper part of the tell.

Middle Bronze IIB (ca. 1800-1700 B.C.). One of the biggest surprises that Tell Qara Quzaq has provided is the fact that during the MB all of the tell, and even its lower part, was occupied by an installation for storing grain, made up of "silos." To date, 42 of these have been excavated. They have walls made of

plastered stone; some are separate, others are in

groups of three, and they have flat floors that include

layers of chaff of varying thickness. Their diameters

vary between 1.8 and 3.2 m and they are 2.5 m in

height, although originally they would have been

higher. There does not seem to have been any urban settlement in the upper part of the tell at this stage of its occupation, and there are only a couple of houses on the northeastern slope. Use of a temple of the earlier level was resumed, possibly for administrative

purposes. Middle Bronze IIA (ca. 1900-1800 B.C.). This level

was destroyed in order to build the silos mentioned above. A temple in antis, excavated in the second season (1990), dates from this period. Its structure underwent a thorough transformation, which suggests a change in its function. It was blocked in by the silos surrounding it, except for the front part, which

permitted access. Abundant indications of what may

24 For previous reports, see G. del Olmo Lete, "Mission arqueol6gica de la Universidad de Barcelona en Siria," Aula Orientalis 7:2 (1989) 269-77; del Olmo Lete et al., "Exca- vaci6nes en Tell Qara Quzaq. Informe provisional: primera campana (1989)," Aula Orientalis 8:1 (1990) 5-13; del Olmo Lete and E. Olavarri Goicoechea, "Excavaci6nes en Tell Qara Quzaq. Informe provisional: segunda campana (1990),"

Aula Orientalis 10:1 (1992, in press); del Olmo Lete and Olavarri Goicoechea, "Excavaci6nes espafiolas en Siria," RArq 11:108 (1990) 29-33; del Olmo Lete and Olavarri Goicoechea, "Excavaci6nes espafiolas en Siria," RArq 12:123 (1991) 52-56; del Olmo Lete and Olavarri Goicoechea, "Tell Qara Quzaq enclave comercial en el reino de Karkemis," RArq 13:135 (1992) 12-15.


have been the environment of this period appear in the materials used in the silos' construction. As far as other structures are concerned, only one wall has been

recovered, in 1991, that was not affected by this con-

struction, as well as two adobe silos, in 1992. In ad-

dition, abundant pottery was found in the external and internal fill of the silos. These remains, typologically similar to those found in other deposits of the

Euphrates valley, such as Tell Hadidi, allow us to establish a reliable sequence for the MB pottery.

Middle Bronze I (ca. 2000-1900 B.C.). There is little evidence from this period for occupation. Ce- ramics were found that cannot be confidently attrib- uted to any structure in particular, but which may belong typologically to this initial phase of the MB era.

Early Bronze IVB (ca. 2200-2000 B.C.). On the

upper platform of the tell, in the level immediately below the area occupied by the MB temple, three houses and the remains of others were uncovered in 1992. They were constructed of crude mudbricks. Their pottery can be dated to the final stages of the third millennium. Although they were massively affected by the construction of the MB silos, they are

large, sturdily built rooms with hearths and water jars. Their walls and floors are covered with a layer of

plaster on top of a layer of reheated earthy paste, following a technique still in use in this region of the

Euphrates. The pottery that accompanies these constructions is very uniform, and is identical to the EB IV forms from Ebla and Tell Hadidi.

Early Bronze III-IVA (ca. 2500-2200 B.C.). During this period, Tell Qara Quzaq was again occupied after a long period of abandonment. The occupation was concentrated in the upper platform. Houses of crude mudbricks were built on stone, and outside of them were silos of stone or mudbrick quite different in manufacture from those from the MBA. One construction of particular importance was found on a

platform, separated from the other buildings by a

pathway flanked to the east by stone slabs set in earth; in the subsoil, a child buried in ajar was found, with the body in the fetal position. The pottery of this level offers a great variety of forms and techniques that indicate a date in the historical horizon of Kurban

Hyiyuk IV and the H phase of Amuq, that is to say,

in the Proto-Dynastic III.

Early Bronze II (ca. 2900-2700 B.C.). The construction named the "red building," which appears in different sectors of the tell, dates from this period. One of its rooms was excavated in 1991. It is a solid

building constructed in baked mudbricks with plastered walls 0.95 m thick. The room mentioned above

(5 x 3 m) was used later as a mortuary chamber, in which two intact corpses were buried. To prepare the room for its function, it was divided into two by means of an adobe wall and the door was enclosed. The

corpses were left in the fetal position with the skull

facing south; the bones were burnt, although there is no trace of ash around them. The wealth of their

accompanying furnishings included seven copper lances, 15 small vessels (a few of them painted), a

large number of beads, and different kinds of fibulae. The pottery indicates that these burials and the build-

ing date from the end of the Amuq G period. The

typological sequence from Tell Qara Quzaq is com-

plete, which allows us to define the evolution of the

pottery of the EB-MB periods. Raqa'i. Hans Curvers, University of Amsterdam,

and Glenn Schwartz, Johns Hopkins University, re-

port: Tell al-Raqa'i is a 0.4-ha early-middle third millen-

nium B.C. site, one of several small specialized communities concerned with grain storage and processing (e.g., 'Atij, Kerma, Ziyada) in the Middle Habur sal-

vage area. A three-week excavation season was conducted by the joint Johns Hopkins/University of Amsterdam team in June 1992 in order to address

questions not resolved by the last full season of excavation in the fall of 1990.25 In addition, two seasons of analysis and small-scale excavation were directed

by Hans Curvers in April and September-October 1991.

A major focus was the early history of the Raqa'i Rounded Building, an edifice some 20 m in diameter that was the central structure at Raqa'i in levels 4 and 3 (equivalent to mid-to-late Ninevite 5 period and

Leilan IIIb-d, perhaps ca. 2800-2600 B.C.). Previous work had tentatively indicated that the structure was founded in level 4, subsequent to the earliest occu-

pation at the site in levels 5-7. To verify this conclusion, a 23 x 1.5 m L-shaped trench was cut through the building down to level 5-7 contexts. For the most

part, the project confirmed architectural discontinuity

25 For prior results, see G. Schwartz and H. Curvers, "Tell al-Raqd'i 1989 and 1990: Further Investigations at a Small Rural Site of Early Urban Northern Mesopotamia," AJA 96 (1992) 397-419, with bibliography. Funding for the 1991

and 1992 fieldwork was provided by, among others, the Dellheim Foundation and Syria Shell Petroleum B.V., Da- mascus.


Fig. 20. Raqd'i. Level 3 enclosure wall with interior buttresses, excavation area 29/120.

between the level 4 Rounded Building and smaller- scale pre-level 4 architecture. The cut also provided evidence of three phases of "grill" architecture in the

pre-level 4 levels northwest of the later Rounded

Building. Also of interest were two more level 4 jar sherds of southern Mesopotamian (ED I?) type, one with a nose lug and crosshatch incision, the other with crosshatch incision only.

A further issue of concern was our limited sample of the earliest levels at the site, 5-7, and the date of this occupation. Therefore, we expanded our exposure of these contexts on the south slope of the tell

(areas 42/114, 42/116, and 60/120). This operation reached virgin soil in several areas and exposed segments of rectilinear architecture. As before, only minimal amounts of pottery or other materials were

retrieved, but these continue to support an early- middle Ninevite 5 date (equivalent to Leilan IIIa or

IIIb) for the site's establishment (e.g., a sherd of a fine gray cup with slightly inverted rim and horizontal

ribbing). Excavation of the fill of vaulted "silo" 2 in the level

4 Rounded Building, partly excavated in 1990, revealed that the room was preserved to almost 4 m in

height. The floor or bottom of the structure was still not reached in 1992, but it would seem unlikely that we were very far above it. Surprisingly, fragments of white lime plaster with traces of black paint were found on the inner face of the southeast wall at a

point that must be near the floor or bottom of the

room. Evidence of wall painting at the bottom of a doorless "silo" might seem to confound an interpretation of the room as a grain storage space, but an alternative explanation is not obvious.

In the northern part of the site (excavation units

29/114 and 29/120), 1991 excavations exposed level 4 domestic architecture. One of the excavated rooms was filled with burned debris, which yielded some two dozen clay tokens. These exposures were extended in 1992 and complemented by a step trench to virgin soil on the north slope of the mound. An unexpected result was the discovery of the foundations of a level 3 (equivalent to late Ninevite 5, ca. 2600 B.C.) type of

community enclosure wall on the northern edge of the tell, sunk into the level 4 houses and extant to a

height of 2.5 m (fig. 20). The wall was some 80 cm wide and included at least two substantial interior buttresses ca. 0.60 x 0.80 m and 0.80 x 1.00 m in area. An interpretation of this architecture as a re-

taining wall (or a set of retaining walls) constructed

contemporaneously with the level 3 houses to the south may be more tenable than a defensive wall, since the structure seems to have been built in awk-

wardly coordinated segments. Preparation of the first volume of the Raqa'i final

reports is underway and includes studies on levels 1- 3 architecture and stratigraphy (G. Schwartz), ceramics (H. Curvers), small finds (S. Dunham), burials (B. Stuart), and archaeobotanical evidence (W. van Zeist).

Among the technical studies being implemented is


neutron activation analysis of Raqa'i ceramics by James Blackman, Conservation Analytical Labora-

tory, Smithsonian Institution, with a view to investi-

gating Middle Habur interrelations and the possible "colonization" of the area in the early third millennium.

Ras Shamra-Ugarit. Marguerite Yon, Maison de

l'Orient, Lyons, reports: The French Archaeological Mission at Ras Shamra-

Ugarit has organized its program into two-year periods, alternating a season of fieldwork with a year of

interpreting the material and preparing the publications. Thus, in 1991, its efforts were dedicated to the

study and publication of previous discoveries and, in

1992, it resumed an important excavation campaign (the 51st), involving investigation of several areas of the tell.26

The present report summarizes the main results of the 1992 excavation, as well as the recent research

developments and their publication in the Ras

Shamra-Ougarit (RSO) series: two volumes (VI, VII) appeared in 1991, two (VIII, IX) in 1992, and one

(X) is foreseen for the beginning of 1993. "Centre de la ville." This excavation area, opened in

1978, relates to the study of architecture at the close of the Late Bronze Age (LBA), and has provided elements of comparison for the understanding of vast zones of the city previously excavated that had not

formerly been published, permitting a study of the

stratigraphy as well as an analysis of the environment. Before ending this program, the excavation of two houses (C and D) had to be completed and certain

points had to be clarified. The excavation of street 1288-1312 toward the

south was an attempt to determine the eastern bound-

ary of the insula and the eastern and western boundaries of the street itself. The walls lie in a disconnected

manner, which causes a widening halfway down the

slope; however, the outline is not always easy to follow because of the many repairs. An interesting aspect of the structure of the street, which is on a rather steep slope, is the presence of small, low, transverse walls

(connected to house walls) supporting the earth of the

slope; furthermore, in the center, a small furrow was

dug, forming a longitudinal drain for running water.

To the west of the street, house C, which completes the northern part of the insula along with houses A and B, is under study for publication. In 1992, excavation of the stairway (with large, monolithic, broken

steps) was finished, as well as that of the southern part of the house.

The Bunker site. In 1973, this point of the city in the sud-centre district yielded a very important series of texts thrown onto the spoil heap following non-ar-

chaeological public works. Here it was possible to resume regular excavation in 1986; the works carried out in 1988 were finished in 1992 by P. Lombard. These three campaigns permitted the complete excavation of the 1973 epigraphic material and its context.

In D 7-8m, we finished clearing the floors down to the foundation in rooms 2050, 2053, and 2072, located west of the ditch 2043 (where the bunker was), and also cleared the old ramp of the bunker toward the south. This operation uncovered some important epigraphic material (see below), as well as a black stone

cylinder bearing a hunting scene (RS 92.2019) and some fragments of Mycenaean pottery that, in several

cases, completed some fragmentary vases found in 1988.

Constructions in D 5m. The excavation still has not reached the boundaries of the building with stone basins previously discovered in this area (1986, 1988), nor has it provided evidence that would clarify the use of the building. Several architectural stages can be observed associated with different floor levels. The excavation also confirmed the fine architectural quality of the building: for example, the systematic use and the careful positioning of the blocks carved at the

angles of the walls, and the modular calibration of the

intermediary rubble. Several new spaces have been opened in the south,

including a room (2093) with four niches placed approximately 0.20 m above the floor. The southeastern

part of the area (D 5m/3) has produced a handsome staircase with a double flight of stairs including the four bottom steps in situ.

"Grand Rue." The urbanism study program--one of the Mission's themes of research since 1978-holds

as one of its goals the precise location of the entrances

26 I thank my colleagues of the Mission-V. Bernard, P. Bordreuil, A. Caubet, O. Callot, Y. Calvet, J. Gachet, B. Geyer, P. Lombard, J. Mallet, V. Matoian, and D. Pardee-- who participated in the publication of the different works mentioned here, and V. Cook, for the English text. For

previous reports, see "Chronique archeologique," Syria 67 (1990) 442-49; Syria (1982, 1983, 1987, 1990); M. Yon, Le

centre de la ville, 38 -44e campagnes (1978-1984) (Ras Shamra-Ougarit III, 1987); M. Yon, "La ville d'Ougarit au

XIIIe s. av. J.-C.," CRAI 1985, 705-21; Y. Calvet and B. Geyer, in Historisches Talsperren II (Stuttgart 1991) 195- 236; Barrages antiques de Syrie (Collection de la Maison de l'Orient mediterraneen 21, Lyons 1992) 69-77.


Fig. 21. Ras Shamra. The Grand Rue area (toward the north): on the left, wall 3008; above on the right, wall 3005.

into the city (the only entrance known until then

having been the western door of the royal palace, which could not have been the main public entrance to the city). Across a north-south depression, which could have been an entrance to the city from the plain beyond the Nahr ed-Delbe, a sounding was begun in 1988. One of the objectives of the 51st campaign was the resumption of this project by V. Matoian.

The excavation in D Im-C Im brought to light an

important thoroughfare (Grand Rue 3016/3018), from 4 to 4.5 m wide, exceptional for urban Ugarit, where the streets are usually much narrower. It climbs from the south to the north between the buildings and seems directed toward the large "square," the eastern part of which had been uncovered more than 30 years ago during the excavation of the "Ville sud"

(in C 3-6 h-j). There, a sounding revealed an important crossroads (fig. 21).

A narrow street (3017), which seems to join the street located in D 5m (Sud-centre district), ended at this wide street. On the other side, a rather vast area with fallen (ashlar) stones seems to have been a public space. Its axis joins one of the streets further to the east in the excavation area of the Ville sud.

This space is bordered on the north by the angle of a building of exceptional architectural quality, whose function is as yet undetermined. The south wall (3005) can already be seen, 1 m thick and con-

structed in stone blocks carefully cut and positioned, with stretchers on edge and headers with interior rubble fill, traces of longitudinal beams, and bedding marks. A fragment of a syllabary tablet (RS 92.3179) discovered in street 3016, and a stone cylinder seal

depicting a horseman (RS 92.3195), apparently came from this building.

This year's excavation has thus confirmed the existence of a major access to the city on the south side,

lying on the axis of the Nahr ed-Delbe, in use during the last phase of the city's history (LB II). This structure under investigation belongs to the urban network that is appearing little by little on the southern slope of the tell, judging from its location in the city, the

importance of the crossroads where large thruways meet, the quality of the neighboring buildings, and

furthermore, the presence of tablets. It will probably provide the missing link for an urban phenomenon that appears in a discontinuous fashion in the other areas of the city.

"Maison aux fours." The work that has been going on for years in the city is beginning to elucidate the last period of Ugarit (the end of the LBA). This period is only perceptible in certain parts of the tell, however, excavated before 1978 and not yet the object of final

publication. This is why we resumed exploration of a zone located near the Palais nord and the Palais royal (in A 15-17 e-f), where the excavation of 1973 had


Fig. 22. Ras Shamra. The bridge-dam pile on the Nahr ed-Delbe, upstream view (toward the east).

started to reveal constructions dating earlier than this last phase, especially the building called the Maison aux fours, which apparently dates to LB I.

In 1992, a sounding of modest dimensions was undertaken without removing any walls, in order to

clarify the present condition of the area. The first observation made was that the house had been excavated at different stratigraphic levels and that the ovens only belong to the final phase.

In the eastern part, the former excavation had

exposed the covering slabs and the top of the access shaft belonging to the family tomb (319). The chamber had been looted, probably during the final phase of occupation, and was found full of saucer lamps in the dromos entry. A fragment of a cuneiform tablet written in Babylonian, unfortunately practically illeg- ible (RS 92.678), was found in the tomb fill, containing boulders fallen from the walls and broken pottery.

In the western room 5, a sounding dug down to a

depth of 1.5 m led to the discovery of four construction levels. Under the two LB I levels, the pavement of a MB II floor (third level) appeared, along with a

jar containing the inhumation of a newborn baby. In the eastern part of the sounding, the top of the wall of the fourth level has appeared. Thus, the sounding provides continuous stratification for the whole LBA.

Hydrography and the bridge-dam. The study of the

hydrographic system by Y. Calvet and B. Geyer en-

tailed the clearance of a pile of the large bridge-dam (D ff-gg 6-7), with a stem allowing for a water reten- tion system on the Nahr ed-Delbe, south of the tell. The hypothesis previously set forth concerning the localization of one of the principal entrances to the

city at the south of the tell has been reinforced by the fact that the Grand Rue mentioned above lies on the extension of this bridge-dam, and could conse-

quently represent the principal access to the urban center.

This exceptional structure has been published in its 1990 state by the two excavators. In 1992, they were able to conduct a more complete excavation that

brought to light new evidence modifying the pro- posed reconstruction. The structure proved to be much more imposing than had been expected. The

part of a massif that appeared on the present left

bank, mostly hidden under a thick layer of sediments, turned out to be a pile in the shape of a symmetrical stem, triangular on the upstream side. The blocks of the downstream side had been torn away (fig. 22). The notches and grooves on it suggest that it was a small beam of the dam, a hypothesis confirmed by the fact that the two channels were paved in order to resist erosion. Thus, it was not a simple construction where two lateral piers framed a single channel, but a complex unit consisting of at least a central pile between two channels, and lateral piers.


Epigraphy. As mentioned above, with the exception of a fragment from the site of the Grand Rue, and another in the Maison aux fours, the important epi- graphical discoveries of 1992 came from the Sud- centre area. Twenty-four complete or fragmentary tablets were discovered and submitted to epigraphers (P. Bordreuil and D. Pardee). These tablets were for the most part covered as usual with limestone deposits (and are now under restoration in the Damascus Mu- seum laboratory), but already it is evident that they are divided between the Akkadian syllabic epigraphy (12 texts) and the Ugaritic alphabetic epigraphy (11 texts); an anepigraphic tablet was also found, bearing only a seal impression.

A mythological fragment (RS 92.2016) is particularly important because it bears part of the colophon of the scribe Ilumilku, already known since 1934 for his copies of mythological texts found on the acropolis. Also remarkable is a new version of a list of divinities (RS 92.2004) reiterating in a nearly com-

plete form a syllabic text formerly published by J. Nougayrol, as well as an incantation (RS 92.2014) against snake and scorpion bites. Several syllabic and

alphabetic epistolary documents and various administrative texts cannot be identified until the restoration

underway has been completed. Sweyhat. Tom Holland, University of Chicago, and

Richard Zettler, University of Pennsylvania, report: The fifth and sixth seasons of excavations were

conducted during the fall of 1991 and 1992. The 1991 expedition was sponsored by the Oriental Insti- tute and the University Museum under the joint di-

rectorship of Holland and Zettler. The 1992 season was directed by Holland and sponsored by the Ori- ental Institute and the National Geographic Society.27

In addition to continuing surface collections in the outer town during 1991, Operation 1 was enlarged, four new operations, 5-8, were begun on the tell, and in the outer town, Operation 4 was expanded and a new operation, 9, was opened. In 1992, work continued in Operation 5 and also in the 1970s area IVN

(=Op. 10) and area X (=Op. 11). Results of 1991 surface survey and Operations 1, 4,

and 9. In keeping with our goal of tracing the extent of occupation in the area surrounding Tell es-Swey- hat's main mound or citadel, surface collections were continued in the area south of the outer rampart. Aerial photographs of the site show a dark line that appears to be a wall abutting the south side of the

outer rampart. Three transects were laid out east to west across the area and 14 units (circles with areas of 25 m2) were placed along each transect. Except for two anomalous areas, artifact (in particular pottery) densities were much lower than outer town scatters, a pattern that T.J. Wilkinson has identified as indic- ative of manuring. In addition, remains of stone architecture were found in one area. Since the pottery diagnostics that could be identified were all late third millennium in date, the area was probably part of the contemporary site, perhaps analogous to the so-called suburbs at Titrish Hdyfik in Turkey. The site would have been close to 50 ha in area at the time of its floruit.

On the main mound, Operation 1 was expanded to the east to include an area that measured roughly 10 x 10 m. At the bottom of the operation, the western wall of a building uncovered in 1989 proved to be the outer wall of a "kitchen," three rooms of which were exposed in 1991. The westernmost room contained an oven, 2 m in diameter, and a circular ash pit; the other two rooms included benches and flat stone surfaces on the floors, perhaps used in food preparation. The "kitchen" was contemporary with the 1970s area IV building (see also Operations 6, 10-11 below). Four occupation levels were uncovered above the "kitchen." The four levels date to the very end of the third and the early part of the second millennium (or an early phase of the MBA), a period that remains largely undocumented. Future work in the area is likely to fill a major gap in the Syrian historical and cultural sequence.

In the lower town, Operation 4, begun in 1989 as a 2 x 2 m test square, was enlarged to more than 200 m'. Seven rooms of a large house and an outside space west of it were found in the area (fig. 23). The house has a floor space of more than 110 m2. Spindle whorls, found on the floors of some of the rooms, indicate that those rooms may have been used for cloth-making activities; a bread oven in the outside space indicates that the area was used for cooking. In Operation 9, excavations yielded parts of three buildings and a street. One of the buildings contained "industrial" installations-shallow ash-plaster basins. A covered stone-lined water channel ran down the center of the street and sloped toward the center of the settlement.

Results of 1991-1992 Operations 5-8 and 10-11. Operation 5 (grid H-J 7), a 10 x 10 m square, was

27 T.A. Holland and R.L. Zettler, "Sweyhat," in Weiss (supra n. 12) 717-19; P.B. Vandiver, M. Fenn, and T.A. Holland, "A Third Millennium B.C. Glazed Quartz Bead

from Tell es-Sweyhat, Syria," in P.B. Vandiver et al. eds., Materials Issues in Art and Archaeology 3: Materials Research Society Symposium Proceedings 267 (1992) 524.


.b.....a. Hearth a HHeart th

bo 0

Operaton 4

Pebbleeinedbb Pit Burnt PPress EPlaster,

0 • ,'Hearth

0. "

ocket

O a 4

Fig. 23. Sweyhat. Plan of Operation 4 building. Shaded remains are from a late phase of operation.

situated north of the 1970s area I to obtain more information about the earlier EB phases of the mound. The upper levels expanded the range of Hellenistic and Roman pottery types, and the latest third-millennium levels produced some unexpected finds. In one of the two pits filled with pottery, found in the southern part of the square, a glazed rock crystal bead was also recovered with an unusually high concentration of copper oxide. The Smithsonian Con- servation Analytical Laboratory reports that "unal- loyed copper, copper oxide or copper carbonate was added to fine quartz powder to make the glaze, which would suggest a link with copper-melting or -using technology or with a malachite technology." This sug-

gestion is strengthened by the 1991 discovery of a stone mold for either a dagger or spear in Operation 5 and another stone mold for making jewelry in Op- eration 6, as well as by two clay crucibles, metal tongs, and other miscellaneous metal objects previously excavated during the 1970s.

The second rare discovery in Operation 5 was the presence of a painted wall fresco in the level preceding the pottery pit phase. Excavation of the fresco and associated deposits continued during 1992; two rooms of a large building containing wall paintings were found, and to date 103 fragments have been lifted and conserved. The first fresco comes from a large wall, 1.25 m wide. Although the fresco is not com-


*Black Maroon

Fig. 24. Sweyhat. Fresco fragment.

pletely excavated, the designs on some of the fragments may be compared to contemporary wall

paintings from Tell Munbaqa and Tell Halawa B, especially the geometric borders, tree branches, and

stylized human figures with "Medusa"-like hairstyles, all painted in bright maroon and black on a white

plaster background. Our fresco, however, contains an unusual new addition from the earlier excavated scenes in that at least one bovine with suckling calf is

depicted standing on a 450 slope (fig. 24), which may represent a mountain, a well-known iconographic feature in Mesopotamian art. The closest parallel for this

type of suckling animal, especially the geometric painted rectangle on the mother's body, is painted on a Late Minoan III sarcophagus from Gournia in Crete, dated to the 14th century B.C. Our fresco may indicate that there were cultural and trade contacts between the Upper Euphrates valley and the Aegean during the Akkadian period; a number of contemporary one-handled storage jars with flat bases from Sweyhat have their closest parallel with an imported and similar jar found in Tomb 164B at Vounous in Cyprus. The second fresco comes from a room to the south of the original painting and only a couple of fragments have been excavated.

Operations 6-8 (grid F-G 5-6) were implemented to excavate more of the large building complex discovered against the western part of the town wall during the 1970s area IV excavations.

Operation 6, a 10-m square located between room 9 and the 1989 Operation 2, resulted in the complete excavation of the central eight-room building unit, 10 x 13.5 m, situated between the town wall and the inner "ring" road of the town. Key finds included metal rings, a stone cosmetic box, and a mother-of-

pearl shell pendant, indicating at least one or more rooms were for domestic use.

Operation 7, 10 m south of Operation 6, revealed the position of at least one town gate through the 3-m wide town wall, which was aligned with a saddle-

shaped depression located on the rampart surround-

ing the outer town in grid G2. The road from the town gate probably leads through the outer rampart, as Wilkinson's geomorphological studies show there is a hollow at that point leading to the Bronze Age settlement of Tell Jouweif in the Euphrates valley floodplain (see below).

Operation 8, located between Operations 6 and 7, was planned to determine the relationship between the central building unit in Operation 6 and the street and gateway in Operation 7. Although only some of the upper EB levels have been excavated, it is clear from the foundation stones of the upper walls that

they rest on top of the ruined wall remains of the

period of Sweyhat's greatest expansion, during the last quarter of the third millennium B.C. Evidence of blocked doorways discovered in the southern wall of the central building unit in the 1970s suggests that there was an adjoining complex of rooms to the south, at least as far as the city gate.

Operation 10 (=IVN, room 8) excavations were renewed in 1992 to complete the plan of the room or

courtyard north of the adjoining central building unit. The remainder of the workbench against the south wall was excavated, revealing another one-handled, flat-based jar similar to the Vounous example cited above. Time only allowed for a small sounding to determine the position of the northern wall (2 m north of the 1970s' north section) of what is most likely a

very large courtyard, now known to measure 6 x 6 m. The most important find from the eastern end of the courtyard was a beautifully modeled clay figurine of a well-defined horse with applied forelock, mane, and sexual organ, indicating the representation of a stallion. A hole bored through the muzzle for a ring to hold reins implies that this is an example of a domesticated horse. On present evidence, this horse model is the earliest known clay example depicting the domesticated horse in the Near East.

Operation 11 (=XA, B-area IVT), a 5 x 10 m trench, was positioned to overlap room 15 and to determine if the rooms built against the town wall continued to the south, which they do. The excavated


portion of the new room, 16, contained many ceramic

storage vessels, a complete bronze dome-headed pin, a vesicular black basalt potter's wheel or pivot stone, a clay mortar, and a limestone elliptical-shaped weight. These finds suggest that room 16 may be associated with a working area, rather than belonging to domestic quarters.

Regional studies: In order to place Tell Sweyhat in its local and regional context, archaeological and geomorphological surveys by T.J. Wilkinson were resumed in 1991 and 1992. On the Pleistocene terrace

surrounding Sweyhat, aggradation was limited to

plow-wash along wadis and to the east of the town wall. In contrast, the Euphrates floodplain, some 3 km to the west, has been extensively reworked by the

migrating Euphrates everywhere except for a small residual of ancient floodplain in the southwest of the

survey area. The Euphrates has truncated Bronze

Age Tell Jouweif, as well as several wine presses at

floodplain level, and it is likely that many sites have been eroded away entirely, thus biasing the archaeo-

logical record considerably. Even allowing for such bias, in this area of marginal

rain-fed cultivation (mean annual rainfall ca. 250

mm), settlement has tended to concentrate along the well-watered floodplain margins where enhanced soil

moisture, if not irrigation, nurtured more reliable

crops. Sites of the Halaf, Ubaid, and Late Chalcolithic

periods were restricted to this zone; then during the

early third millennium, small settlements extended across the terrace around Sweyhat. The early phase of Tell Sweyhat, which probably occupied 5 ha or less,

belonged to this phase of expansion. Linear hollows

running from southwest to northeast suggest that

Sweyhat developed on an ancient route leading to the northeast from the Euphrates valley. When Tell Swey- hat attained its maximum size during the final quarter of the third millennium, land use intensity, as suggested by extensive sherd scatters across the terrace surface, also peaked. There followed a contraction of settlement during the MBA so that by the LBA, settlement had retreated back to the floodplain margins. Occupation of the terrace in the form of small dis-

persed settlements resumed during the Iron Age and

continued through the Hellenistic, Roman, Byzan- tine, and Early Islamic periods. There was again a retreat back to the floodplain by the 12th/13th centuries A.D. Although it is unclear whether these ex-

pansions and contractions were climatically or

culturally driven, suggestions that the ancient environment was more favorable come from the presence of several probably Hellenistic to Late Roman wine

presses cut into the limestone at the base of the Eu-

phrates bluffs. Significantly, no grapes are grown in the region today.

SECOND MILLENNIUM B.C.

Ashara-Terqa. Olivier Rouault, Collkge de France,

reports: For the last two seasons of excavation, the principal

objective of excavation concerned the "post-Mari" period, for which the site of Terqa is particularly im-

portant; however, in accomplishing this task, we have

gleaned new information on other periods, in particular the period contemporary with the Mari archives, and the Islamic period.28 Furthermore, having com-

pleted the study of extensive second-millennium levels, we began to excavate older levels.

Fieldwork was mainly conducted in areas E and F. Islamic levels were uncovered to the south of sondage SG40 (TPR 10). Preliminary analysis of the ceramics indicates that the Islamic occupation reemerged in the Umayyad period and continued until the Ottoman

Empire. Under this phase of occupation, a large official building was uncovered that yielded a cuneiform archive attesting to the maintenance or restoration of

Babylonian control under Ammisaduqa, followed by Samsuditana, the sovereigns of Hana unknown at that time (Kuwari, Zimri-Lim, and Hanaya for example), and the seat of Mitanni power for the region, under the reign of Paratarna and Saustatar.

The official nature of this building may be confirmed by the retrieved door sealings, and also by the

discovery-besides the tablets-of a complete bronze

harp (length 59 cm, thickness 0.7 cm) in an excellent state of preservation.

In sector F, the clearing of a large architectural structure continued. In light of the diverse finds (round tablets and lists, door sealings), it appears to have been an administrative or scholarly building. Most of the material is contemporary with the Mari archives, but the levels below this phase have yielded documents of shakkanakku type.

The Old Babylonian architectural structures showed great unity in construction methods and di-

28 The Institut frangais d'archeologie du Proche-Orient has aided us through the services of Yves Baudoin for top- ographic measurements and aerial photographs. The fieldwork was financed by many institutions: the DGRCST of the Ministere des affaires etrangeres frangais (G. Buccellati), and

the Association of Meeting per l'amicizia fra i popoli (A. Smurro). The field photographs were subsidized in 1991 by Kodak-France, and the 1992 expedition was facilitated through the support of Elf-Aquitaine.


Ity,

tAi

r..

Fig. 25. Bi'a. Mosaic 2 in Byzantine cloister.

mensions. These structures, perhaps in function at the beginning of the Hana period, must be seen as

contemporary with the Mari archives. Many phases of construction are apparent. It seems that the remains of a first building-dating from the beginning of the Mari period (Yahdun-Lim)-served as the foundations for a later building (Samsi-Addu or

Zimri-Lim). The oldest structures are arranged according to a

very complex plan, with many successive reconstructions preserved on a few beds of brick. A sounding dug in the 14th season revealed a structure of baked mudbricks (brick size 34.5 x 3 x 10 cm) approximately 2 x 2 m, with a floor of baked brick and

pebbles. A second room was revealed to the north, but it is covered by the balk, and study of this will be a priority for the next season.

The excavations at Terqa have yielded an interest-

ing collection of figurines: in sector E two "Ishtar"- centaur figures were found, of which there are no other examples known from the Near East in the second millennium. The front part is molded on the model of a woman holding her breasts in her hands, and the back is in the form of a quadruped.

Finally, a preliminary study of the seals and seal

impressions from sealings and tablets (certain tablets

carry the traces of 18 impressions) has been under-

taken to understand this important documentation, which includes many royal seals. Different styles have

appeared, some even on the same tablet: Old Baby- lonian, Hana, Kassite, Anatolian, and Mitanni. The oldest impression, found in the deepest levels of sector

F, shows ED characteristics, but could be a copy made at the end of the third millennium or beginning of the second millennium. The presence of a number of these impressions on dated tablets, or those mention-

ing the name of a sovereign, are of particular interest.

Bi'a. Eva Strommenger, Museum fir Vor- und

Friihgeschichte, Berlin, reports: Excavations at Tell Bi'a continued from the middle

of August through the end of October 1992. Our efforts concentrated on the Central Mound E with the following results:

1) The three small eastern church rooms of the

Byzantine monastery were cleared and a beautifully colored, well-preserved mosaic floor was revealed in the middle room (fig. 25). The rooms north of the church held several small finds, as well as interesting architectural features, particularly the refectory, which had omega-shaped benches (fig. 26). We plan to continue and complete the excavation of this room

during the next season.

2) Under the monastery are four or five levels of mudbrick walls dating back to the first half of the


Fig. 26. Bi"a. Refectory with benches.

Fig. 27. Bi'a. Akkadian administrative building and part of Early Dynastic "Old Palace" (left).

second millennium. Seen in conjunction with the previous seasons' excavations, these structures seem to indicate private houses rather than public buildings.

3) Sixty economic texts from the last years of the

reign of Shamsi-Adad I were found at the northern

entrance to the "Young/New Palace" at the late Old Babylonian level. One of the larger tablets mentions the city name "Tuttul" twice. This tablet records a delivery, ending: "this delivery was received in/at Tut- tul." Thus, Tell Bi'a can now be firmly established as ancient Tuttul. The rooms to the west and to the east of the entrance will be excavated next season as they may yield further archival material.

4) Clearing the late ED "Eblaite" "Old Palace" walls revealed a doorway leading north from room 1, and

along the western outer wall a stone drain coming from the northern bathroom was exposed (fig. 27). At the same time, the outer wall of a large public building appeared immediately to the west with a columned facade, dating to the Akkadian period, previously buried under brick collapse from the earlier palaces. Clay sealings confirm the date of this building, rich in finds. Large ovens used for firing clay figurines and ceramics were exposed.

5) A limited excavation of the pre-"Old Palace" building (already begun during the 1990 season) yielded large amounts of ceramics and valuable jewelry, including a gold toggle pin, and a piece of fur- niture with inlays.

6) Work in the deep trench had to be terminated before we reached virgin soil. The lowest levels were already below the groundwater table of the immediately surrounding fields. The levels exposed date to an early phase of ED.


Fig. 28. Hammam al-Turkman. Part of drainage installation in Middle Bronze I palace, facing northwest. Diameter of clay "pipe" is 57 cm.

7) Finally, we reexamined the private houses on Mound B South where they join the city wall, which served to complete the outline of the city plan.

Hammam al-Turkman. Diederik Meijer, Univer- sity of Leiden, reports:

The sixth campaign at this large site was held from May through July 1992.29 The twofold goal of the excavation project is the elucidation of the regional occupational history, on the one hand, and in-depth knowledge of the Middle Bronze Age (2000-1600 B.C., called period VII in the Hammam sequence), on the other. In 1992, most of our efforts were directed at the latter period; both on the summit and on the southern slopes, the occupational remains were

extensively investigated. On the summit, the excavation of the MB palace

complex was continued. The building history of a rather enigmatic square chamber dug into the "official" wing of the palace has been clarified. Although it had earlier already proved to be empty, its function is now thought to have been that of an (unfinished) tomb for one of the rulers; unfortunately, there is no way to establish the latter's identity. The complex itself proved to be quite monumental in its earlier phases. Thus, the southeastern wing, which during later times had been adapted to domestic features such as hearths and ovens, originally had an audience hall with a double doorway and wide, well-plastered walls. A

quite large drain in the form of superimposed clay "pipes" had been installed in several places in the

building, in one case accessible via steps (fig. 28). Sealings and pottery indicate a date of around 1850

Fig. 29. Hammam al-Turkman. Alabaster house altar from the Parthian (?) period. Height 25.5 cm.

B.C. for this phase, and earlier phases are still to be excavated. A few business texts were found in the debris associated with rebuilding activities. Although several insignificant interruptions in the occupation of the complex have been noted, it must have been in virtually constant use from ca. 1900 B.C. or earlier until the end of the MBA. Subsequently, the same functions in the community were fulfilled by a LBA palace (HMM period VIII) built on the same spot.

Under remains from the Parthian/Roman period (period X) that are at least 1 m thick, MBA remains also came to light on the southern slope of the mound. From the configuration of the tell and from small soundings in the intermediate squares, it may be sur- mised that there was no structural connection between the MB remains in the south and those on the main summit. The resulting hypothesis of a dual settlement in the MBA is at present one of the main foci of our work. Although sounded in 1988 during the fifth campaign, the MB levels on the slope were now investigated in two excavation squares. The careful removal of the thick accumulation of the later

occupation, which yielded such interesting objects as a "Parthian" house altar (fig. 29), unfortunately left

29 The sixth campaign was financed by the University of Leiden and the Netherlands Organization for Scientific Re- search (N.W.O.). Earlier reports can be found in Weiss (su-

pra n. 12) 719; Akkadica 64/65 (1990) 1-12; and M.N. van Loon ed., Tell Hammam et-Turkman I (Leiden 1988).


Fig. 30. Afis. Area E. Late Chalcolithic wall, M. 1155, and the Middle Bronze II wall, M. 1115, from the west.

little time for work on the Middle Bronze layers proper. However, houses with ovens and well-plastered walls and floors are attested, next to the burials found there earlier.

The architecture from the late period (period X)

overlying the MB vestiges in the south confirms our earlier description of similarly dated structures on the main mound: the (often thick) walls are dug deeply into the subsoil-thus destroying a good part of earlier remains-and consist of poor-quality bricks laid with large seams between them, with low-quality mortar being used. The whole situation gives the impression of quick work, not overseen by a master

bricklayer.

FIRST MILLENNIUM B.C.

Afis. Stephania Mazzoni, Universita di Pisa, re-

ports: In 1991 and 1992 progress was made in two differ-

ent areas of the acropolis: area E1-2 on the western

slope, where already in 1988 and 1989 a long se-

quence of Iron Age I levels had been clarified, and area G on the eastern side, near the tomb of the Sheikh Hasan and a small modern cemetery, where in 1989 a large rubbish dump of Iron Age II-III and the remains of the settlement of the late Iron III

(Neo-Babylonian/early Persian period) were excavated.30

30 For previous reports, see S. Mazzoni, "Afis," in Weiss (supra n. 12) 729-32; Mazzoni, "Tell Afis and the Chronol- ogy of the Iron Age in Syria," AAS 40 (1990) 2-4; Mazzoni, "Tell Afis e il Ferro I in Siria," in Mazzoni ed., Tell Afis e l'etit del Ferro (Pisa 1992) 157-67. The 1992 campaign was undertaken by S. Mazzoni (director), S.M. Cecchini (codi- rector, area G), P. D'Amore (area L), G. Scandone Matthiae

(filing objects), B. Wilkens (palaeozoologist), L. Lazzarini and R. Falcone (geologists: clay, stone, and pottery sampling), G.M. Ingo and S. Fontana (chemists: iron and bronze sampling), D. Bonatz, M. Degli Esposti, and F. Venturi (archaeologists), M. Necci (photographer), L. Scardala De Ninno, S. Martelli, and M. Epifani (draftsmen), and A. Graziani and G. Buonomini (restorers).


Fig. 31. Afis. Area E. Cylinder seal with linear pattern.

Three important results have been obtained in the last campaigns. The first is the discovery in area E1 of a massive city wall dating to the final stage of the Late Chalcolithic period and sealing a long sequence of levels of the earlier settlement going back to the Ubaid

period. Noteworthy is the discovery of the mudbrick defensive wall of the citadel from the MB II period, overlying the EB III-IV settlement and the earlier wall. Finally, at a very deep level and under a dense

sequence of Iron Age I and II private units in area

G, we cleared out the facades of a ceremonial building enclosing a large pebbled square, to be dated between LB II and early Iron Age I.

The Late Chalcolithic levels. In area E1 the slope trench was extended westward, reaching a total length of 30 m and exposing a succession of 26 levels from the top of the acropolis to its base in the western Lower City. This base is clearly built up by the two massive and well-preserved walls of the Late Chalco- lithic and MB II periods strictly superimposed, with the more fragmentary remains of EB IVA-B in between (fig. 30). The wall of the Late Chalcolithic phase (M. 1155) is a 2-m high solid structure in stone with an outer face of boulders laid in three slightly sloping terraces; over this base lies a 1-m high superstructure formed by levels of pebbles alternating with a mixture of clay mud and a few bricks over straw mats treated with a sort of whitish plaster for waterproofing. The

upper face of the upper course of the stone-built base,

exposed in a small central section, is made up of a

filling of pebbles serving as a real terrace for this

superstructure. The floor in front of the wall sloped slightly down toward the Lower City; probably a shallow ditch faced the wall, ensuring the runoff of rain for both water collection and prevention of any dam-

age to the structure. A thick homogeneous deposit covered the wall,

probably laid in different strata but containing pottery and flint of the same cultural horizon; only in the debris filling the very top of the wall were a few artifacts and sherds collected, witnessing a later date

in the Jemdet Nasr period, namely, a fragment of a

jar painted in a stylized animalistic style of early Nine- vite 5 with a potter's mark, a limestone cylinder seal with a linear pattern (fig. 31), and a miniature animal shell inlay.

The ceramic horizon of the wall's fill is well characterized by the coexistence of painted Late Syrian Ubaid pottery in the forms of bowls and cups with

simple geometric patterns decorated in a brownish- blackish or reddish thin paint, and everted rim jars and Chaff-faced ware in the prevalent form of Coba

bowls, mostly flint-scraped. The lithic industry consists mainly of obsidian blades, triangular microliths, and Canaanean blades.

This horizon is well represented across a large geographical area covering south-central Anatolia (Mal- atya VII, Kurban Hbytik), southern Anatolia (Coba

Hytiuk/Sakcagozu), northeastern Syria (Tell Leilan, Tell Brak), northern Syria (Tell Hammam al-Turk-

man), coastal Syria (Amuq F, Ras Shamra III 2), as well as central Syria (Hama, Qal'at el Mudiq).

The presence of administrative buildings (Malatya VII, Tell Hammam VA) and even of massive city walls

(Afis,Jawa, Sidon, Ras Shamra, Byblos) for this period is a clear document of the intensity and diffusion of a local process of proto-urbanization and the formation of complex societies, on the eve of the Uruk

expansion or at its very beginning. In Afis, the same long duration of this local phase

is documented by the archaeological evidence obtained in the sounding at the base of the wall; here, at least three distinct phases of occupation are represented from levels 19 to 25, the last level consisting of an open paved courtyard with a large funnel-

shaped hearth. The presence in these levels of a large storage jar in coarse ware and a fragment of a Re- served Slip jar together with Chaff-faced and Late Ubaid painted wares indicates a possible correlation, at least for the central part of this local sequence, with the Middle Uruk phase of Mesopotamia. Finally, in a

very small section of this sounding, a deep burnt fill


of level 26 was cleared with large quantities of materials dated to the Late Ubaid period. More than 6 m of archaeological accumulations separate this level from the virgin soil, as we know from a perforation made at the end of the season at the western edge of the sounding; by comparison with the long sequences of the Halaf-Ubaid periods documented all over northern Syria, it is possible to assign this deep fill to these phases. As the virgin soil is at least 3 m deep over the surrounding plain outside the Lower City, the earlier settlement of Afis had to be founded over a natural limestone spur in a position of clear visibility of the northern alluvial plain with its natural passages to the coast, the northern Taurus, and the Euphrates plain.

The Early and Middle Bronze Age levels. After a long gap in EB II and III, the area was newly occupied in EB IVAl and EB IVA. To EB IVB we date a large mudbrick wall running along the earlier mound, possibly the remains of a defensive wall. Inside, two rooms were identified that contained permanent in-

stallations, a plastered platform, a clay vat, a potter's wheel, and a large pottery assemblage of painted and

plain wares. In MB II the area was leveled for laying out a large defensive wall (M. 1115). The foundations consisted of one course of large blocks with an upper fill of small stones, pebbles, and sherds as a base for the mudbrick solid structure. To the outside, over the Lower City to the west, a gentle slope of stones protected the base of the wall. Small corridors were

opened to both the outside and the inside of the wall,

giving it the appearance of a casemate structure; the

southern, and innermost, corridor was used as a stor-

age room. Against the inner side of the wall a storage unit was uncovered, containing in situ jars and vats dated to MB II.

The Late Bronze II and Iron I levels. The area was

occupied in the LBA; only the thick uppermost levels

overlying the MB II settlement have been excavated,

dating to the final part of LB II and immediately covered and partly reused by the Iron I structures. The first Iron I settlement was laid directly upon, and

thereby partially readapting, the earlier structures; its date can be firmly fixed to the second half of the 12th

century on the basis of a Submycenaean bowl, painted with spirals, which is the example documented fur- thest inland in Syria up to now. A large room, probably a courtyard, has been excavated for these phases, surrounded on three sides by large walls; on its western side over the slope, in a storage room, a small deposit of well-preserved pottery (storage jars, bowls, dishes in Simple ware) was brought to light, providing good documentation for the pottery horizon of this phase. The pottery tradition is characterized by a clear

continuity from the LB II phase; the biconical and ovoid storage jars with combing on the shoulders were still in fashion, but the cigar-shaped elongated jars with brilliant but irregular burnishing were by this time prevalent. Characteristic of the local pottery horizon is a painted ware decorated with simple geometric patterns, reminiscent of the Submycenaean, Cypriot, and Levantine painted wares of the final LB tradition.

No cultural break can be detected between LB II and Iron I in the material culture, a sign of the

continuity of the cultural development in this region. There is, in any case, an internal transformation in the nature of the settlement: some domestic and rural structures, an open area with hearths, and a mudbrick silo replaced the earlier buildings, witnessing a shift in economic and social conditions. These data support the picture offered by the historical documents of a

region undergoing a limited decline in urbanism in connection with the sedentarization of the Aramaean

tribes; they reject, on the contrary, the picture of Iron I as a "Dark Age," marked by a clear break with the earlier local cultural traditions. During Iron I, in fact, the area underwent many adaptations from a rural

implantation (Iron IA: levels 9a-c and 8) to a clearly well planned settlement (Iron IB: levels 7-6), with a trend toward greater organization and specialization of units (Iron IC: levels 5-3), reflecting a growing stability for this center up to the ninth century.

Area G. The same long and continuous sequence of the Iron I levels was obtained in the eastern acropolis, where the excavations extended over an area of more than 100 m2 and locally to a depth of nearly 6

m. Here, small sectors of a public unit came to light, consisting of a large pebbled square with a north- south length of over 11 m and an east-west width of 6.5 m in the area thus far exposed. It was bordered on its eastern side by a mudbrick wall, 2 m thick and

preserved to a varying height of 1-2 m, possibly the

long facade of a ceremonial building, and on its southern side by a second, better-preserved wall, crossing the first one; at the junction of these, the inner elevation was apparently strengthened by a circular solid

superstructure of bricks. The original elevation of the eastern front wall certainly reached 8 m, as its collapse covered the square regularly and almost horizontally. The pebbled floor was joined not to the base of the walls, but to their elevation, as is clear from the fact that the bricks of the walls near the floor were only partially in sight; thus, the pebbled floor was not the original floor of the square.

The square was in use at the beginning of Iron Age II, when a shallow fill accumulated over the pebbled pavement, covering also a portion of the walls. After


the collapse of the walls (which probably resulted in a vast depression), the area was used as a rubbish

dump with a special function, and slowly filled up with hundreds of animal bones and large quantities of sherds of fine fabric, mostly Red Slip with a few Western imports, dating to Iron Age II and III

(eighth-seventh centuries B.C.). A general rearrangement of the area took place after the collapse of the walls bordering the square, and the buildings with their long front walls were cut and leveled and finally resettled. Five building phases followed one another in Iron Age I with a dense sequence of occupation paralleling at least one part, possibly the latest, of the

sequence of area El and sharing with it the pottery horizon. This phase is marked by the diffusion of the

painted ware and the persistence of the earlier forms of the storage vessels, and clearly predates the intro- duction of Red Slip ware. The buildings were used for storage and domestic activities, as proven by the in situ presence of large jars, tannurs, and hearths. A final rearrangement took place during Iron Age II- III. To these phases date some poor walls mostly in

pis6, preserved under the surface. The date of the square and the facades can only be

inferred. The terminus ante quem of the refilling of the pebbled floor, containing materials of the Iron

Age II, suggests, at least for the final use of this floor, some date during Iron Age I. The long occupation and transformation of the peripheral areas during Iron Age I pushes back the date of this unit to the earlier part of Iron Age I. As mentioned above, the

pebbled floor is certainly not the original floor of this unit. Our current hypothesis is that the original laying out of this complex dates to the final LB II, followed

by a continuity of use and transformation through the earlier part of Iron Age I, which would be consistent with the continuity of occupation revealed in area El.

To the same chronological span belongs a pebbled road brought to light in area L (excavated in 1992), to the south of area G, slightly rising to the acropolis and probably connecting it to the Lower City. Here too, immediately beneath a thin accumulation of Iron

Age II-III, the excavations revealed a longer and dense sequence of levels of Iron Age I, testifying to the extension and stability of the occupation for this phase.

These results lead us to correct partially the picture of the main urban growth of this site through Iron

Age II-III, ninth-seventh centuries B.C.; the long sequence of levels of Iron Age IA-C pushes back this process at least to a period between the 11th and the

beginning of the ninth century, i.e., to the phase of the formation of the Aramaean kingdoms and the consolidation of the Luvian ones. Which of these were

responsible for this cannot yet be answered. In any case, the Iron Age culture emerged gradually from its LBA local background.

Ahmar. Guy Bunnens, University of Melbourne,

reports: Two more excavation campaigns were conducted

at Tell Ahmar in 1991 and 1992 by the University of Melbourne.31 Three areas were investigated. Work continued in the Neo-Assyrian building of area C at the west end of the site. A new area, E, was selected for further investigation in the northern part of the site, and a stratigraphic sounding, area S, was started on the eastern slope of the acropolis.

Area C. Four rooms on the northeast and northwest sides of the courtyard of a Neo-Assyrian building had been identified and partially excavated in 1989 and 1990 (fig. 32). More work was done in this part of the

building in order to expose the original floor, and the excavation was extended to the south in order to

recognize the plan of the building. An area about 40 x 40 m has been exposed but only the northeast, northwest, and southwest boundaries of the building have been recognized. The southeast end must lie further southeast, in the direction of the modern

cemetery of Tell Ahmar. The extension of the excavated area revealed the

existence of three other constructions. One, called building C2, stood against the northwest exterior wall of the building excavated since 1989 (henceforth referred to as building Cl); another building, C3, stood against the west corner of building Cl, and a third

building, C4, had been erected against the southwest wall of building C1. Parts of two rooms of building C2 and only a corner of both buildings C3 and C4 have been exposed. This is enough, however, to show that no preconceived plan seems to have structured

occupation of this part of the site in Neo-Assyrian times.

The main features of building C l can be described as follows: most of the excavated rooms were distrib- uted around courtyard III, which was about 25 x 10.5 m. Another courtyard, marked XIII on the plan, probably existed to the southeast, and another series

31 The 1991 season was sponsored by a grant from the Department of Classical and Near Eastern Studies of the University of Melbourne. In 1992, financial support was provided by the Arts faculty of the University of Melbourne,

and the Department of Classical and Near Eastern Studies of the University of Melbourne. Special thanks are due to LapPaq Computer (Australia), Pty Ltd., and Sokkisha Pty Ltd. (Melbourne).


II

3i I -

. .... Xll

\ ,IV

VI XI

VII

VIII IX

XIII

0 5 10.m knolL ?---. ?

Fig. 32. Ahmar. Reconstructed sketch plan of archaeological structures in area C, after 1989-1992 seasons.

of rooms may have been disposed around it as well. Room X might have been a reception room between the outer and inner parts of the house. Such a structure, with two courtyards, would correspond to the standard plan of Neo-Assyrian houses.

The only entrance identified so far was near the west corner, at the southwest end of the northwest exterior wall. A lane paved with baked bricks gave access to it. The most remarkable feature of the entrance complex, however, was the mosaic-like pavement that lay on both sides of the access lane. It was made of black and white river pebbles disposed in a checkered pattern. A similar mosaic pavement was also unearthed in area E at Tell Ahmar and, further north, in the "baitiment aux ivoires" at Arslan Tash.

It is possible that an upper story existed at least over the northeastern part of the building. A flight of stairs was built at the northeastern extremity of the southeast wall of courtyard III. It could have led to an upper floor rather than to the roof.

A large number of complete or reconstructible pots and hundreds of sherds were recovered. They include

Red-Slip ware, the so-called Palace ware, and a fine

gray ware imitating basalt implements. Although building Cl appears to have been system-

atically stripped before its destruction, a nice female ivory head, three cylinder seals (one is illustrated in

fig. 33), another stamp seal, and many bronze, iron, bone, and stone objects were recovered during the 1991-1992 excavations.

Area E. As part of a series of soundings planned in the Lower City, an area was selected for excavation in the northern part of the settlement. About 20 x 8.5 m in size, it revealed the walls of another Neo-Assyrian building immediately underneath the surface. A

courtyard, more than 14 m long, was bordered to the west by a long wall with a doorway in its middle. The

courtyard itself was covered with a mosaic pavement of black and white river pebbles disposed in a checkered pattern similar to that of the entrance to building C1 in area C. To the south of the doorway, on the other side of the wall, was a recessed niche with a slab covering its floor and tiles coated with bitumen pro- tecting the base of its walls. In front of the niche, but


Fig. 33. Ahmar. Neo-Assyrian cylinder seal from building C1.

not exactly on the same axis, was a quadrangular platform made of baked bricks. This device is strongly reminiscent of the reception rooms of Neo-Assyrian houses. The walls of the room were coated with several layers of white plaster. In the tumble to the west of the long wall, pieces of paintings, essentially black concentric circles and straight lines on a white back-

ground, were discovered. This indicates that the upper part of the wall must have been decorated with

geometric designs. Apart from a clay cylinder seal, two bronze fibulae,

and a few sherds, no significant finds were made in

this building. Area S. A site to the northeast of the deep sounding

of the Thureau-Dangin excavations, on the acropolis, was chosen to start a step-trench aimed at providing a stratigraphic sequence of the occupation periods attested at the site. Iron Age levels still exist in this

part of the acropolis, so it is likely that we can get a

complete sequence from the early first millennium B.C. down to the earliest levels of occupation.

Work started in 1991, but was interrupted in 1992. It will be resumed in 1993. The first excavation season

in area S uncovered a large and complex wall structure. A sort of casemate wall, still about 2 m high, stood on stone foundations. The width of the wall is not yet known but seems to exceed 8 m. Unfortu-

nately, few sherds were recovered and thus the wall cannot be easily dated. A date in the Iron Age is, however, likely. This would show that, in the first millennium B.C., the Til Barsip acropolis was fortified. Apart from a few sherds, more than 80 pieces of iron were discovered, often very corroded. Several of these pieces were armor scales-not an unexpected find in a fortification wall.

HELLENISTIC, ROMAN, AND ISLAMIC

Bosra esh-Sham. J.-M. Dentzer, Universite de Paris, reports:

Since 1981, in collaboration with the Directorate of

Antiquities of the Syrian Arab Republic, the ERA 20

(Centre de recherches archeologiques, CNRS) has been engaged in the study of the urban development of Bosra, both the organization of the city itself and the rural areas around it.32 This program is attempt- ing to establish the relationships among at least four

32 For previous reports, see J.-M. Dentzer, "Les sondages de l'Arc Nabatden et l'urbanisme de Bosra," CRAI 1986, 62- 87; Dentzer et al., "Sondages prbs de l'Arc Nabatden de Bosra," Berytus 32 (1986) 163-74; Dentzer, "Ceramiques et environnement naturel: La ceramique nabatdenne de Bosra," in Studies in the History and Archaeology of Jordan II

(Amman 1985) 149-54; R. Al-Mukdad and Dentzer, "Les fouilles franco-syriennes a Bosra (1981-1987)," AAS 37-38 (1987-1988) 224-41; Dentzer, "Fouilles franco-syriennes Ai l'est de l'Arc Nabatden (1985-1987): Une nouvelle cath&- drale ai Bosra?" XXXV Corso di cultura sull'arte ravennate e bizantina (1988) 13-34; Dentzer, Bosra, Contribution fran- Caise

N l'arch'ologie syrienne (1969-1989) (Damascus 1989)

147-49; Dentzer, "Neue Ausgrabungen in Si'(Qanawat) und

Bosra (1985-1987): Zwei einheimische Heiligtiimer in der vorkaiserzeitlichen Periode," in Akten des XIII Internationa- len Kongresses fiir klassische Archdologie (Berlin 1990) 364- 70; R. al-Mougdad et al., "Un amphithedtre a Bosra?" Syria 67 (1990) 201-204; H. Broise, "Vitrages et volents des fe- netres thermales Ai l'epoque imperiale: Les thermes ro- mains," in Actes de la table-ronde organisee par l'Ecolefrangaise de Rome (Rome 1991) 61-78; Dentzer and P.-M. Blanc, "Techniques de construction et de revetement dans la Bosra nabat'enne," V eme Congres sur l'histoire et l'archeologie de la

Jordanie, University of Science and Technology, 12-17 avril 1992 (in press); Dentzer, Siedlungen und ihre Kirchen in

Siidsyrien, Exhibition catalogue, in press.


m9

Pibce Q

Passage Q PlleMord

secteur 7

seclteur 8 -Passage N

m 67 IY FF 10

F-4 /F 5 m 67 Fl I

42 --'F

11

F 10

NABA5TEE Secteur 5

F, F 2 5 ---- --- A-J

,, Q-,- Mi• x aCir~r

1 31

m- PorteP Passage S

.

F 30 F 35' secteur 6

m i i I

HASATEEN scer5d1 moself xd 15 UY*( )(d 38

mM BSSgeort Fm 12 r7 Pasag coup*

(lt2315m F

Fig. 34. Bosra. Chantier A with Nabataean Arch.

types of land divisions that have been identified on

early aerial photographs, and the areas of the city, two of which have the same orientation as surround-

ing fields (J. Leblanc, M. Dodinet, J.-P. Vallat, F. Villeneuve). The relative chronology and absolute

dating evidence are also being investigated. The Nabataean Arch (P.-M. Blanc, J.-M. Dentzer,

J. Dentzer-Feydy, P. Tondon). The complex structure called the Nabataean Arch, which is at the east end of the main east-west street of the city, provides a key to the interpretation of one of the major phases of urban development (fig. 34). The facade of the actual

arch, on the west, is more or less perpendicular to the main street that constitutes an axis for the western

part of the town. To the east this arch itself is linked, via two oblique walls, to two broad piers, also decorated with Nabataean half-columns and imposed on a different alignment. This break in axis corresponds to a difference in orientation observed between the central part of Bosra, to the west, and the sector situated to the east that has already been identified as the "Nabataean district." The whole structure consti- tuted by the arch and the two Nabataean piers can be

explained as a hinge between the new quarter and the older city center. Its dating provides a chronological clue for the monumental development of the city toward the east under the impulse of Nabataean political power.

Additional soundings carried out on the arch in

May 1992 not only revealed traces of earlier occupation in the area (pottery from the EBA to the Helle- nistic period, including Iron Age material), but also clarified the building sequence. When work started,

probably around the middle of the first century A.D., the project only involved the arch itself, simply com-

pleted to the south and north by porticoes. During construction of the elevation, this project was modified. An extension toward the east was planned: for this purpose, blocks built into the east facade of the arch waited to be used as voussoirs. At the south corner of this facade, a first wall, roughly perpendicular to it, was perhaps partially built following this

plan, but then very likely a major modification of the

project involving the whole district occurred. It resulted in a different, and quite oblique, orientation for the two walls. At this time, the waiting blocks were broken or recut. The final construction, dated to before the end of the first century, delimits a passage onto which open a room (P) to the south and three

bays (N, Q, and S), one (Q) serving an annex area to the northwest, and the others (N and S), a portico with Doric columns on either side of the masonry piers with Nabataean half-columns.

The east sector (P.-M. Blanc, J.-M. Dentzer,

J. Dentzer-Feydy, P. Tondon). The large monument that could be the cause of these changes was discov-


iA

if ,o X0

K ..

N U tl

,

O 10 50m 2 lil~ c= L ~

Fig. 35. Bosra. South Baths.

ered in an extension of the excavations toward the east. It was a vast enclosure, comparable to the temenoi well known in the Graeco-Roman sanctuaries of the

region. It was discovered under two monumental,

superimposed constructions. The last is a large church of central plan dating to the end of the fifth century A.D. The enclosure wall, followed over 25 m, is decorated with pilasters and covered by plaster moldings. It was preceded by a portico with columns covered by plaques of fine-grained pink limestone, on some of which graffiti are preserved. The only feature preserved within the central area is part of a leveled-off wall foundation that could be the corner of a structure. The pottery and coinage recovered suggest a date for the monument toward the last quarter of the first century A.D.

The South Baths (H. Broise, V. Jolivet). On the site of the South Baths, the only traces of earlier occupation are a few sherds (Eastern sigillata, black-glazed wares, Nabataean painted and coarse wares) and two

perpendicular walls, one of which was completely leveled off. Numerous reused blocks with Nabataean letters make one suspect the occurrence, in this area, of one or more monumental structures of the Naba- taean period.

The original thermal building is small but monumental in character, which implies that it was intended to be extended later (fig. 35). The building's plan, designed in the second century A.D. on a right- handed circular layout (rooms S, L, M, N, K), was

rapidly modified. Thus, in the third century, the addition of a second caldarium (U) and corresponding

service areas (W), as well as the creation of two tepi- daria (J and T), gave the building a plan similar to imperial baths with a symmetrical arrangement of warm rooms. In the fourth century, transformation of the octagonal frigidarium into a tepid natatio dou- bled the number of cold rooms with the creation of the two frigidaria S and DF as well as the two adjoining apodyteria. The two palaestrae C and X were probably created in the same period.

The monument thus attained its maximum size and from then on only underwent internal modifications before being gradually used for other purposes. At a date that is not yet clearly determined, a monument with apses, very probably a church, was built to the detriment of the northern part of the east palaestra. Constructed on a higher level than the Baths, this monument overlies a series of shops opening onto the decumanus and a covered passage allowing direct access to the palaestra from the street.

Finally, in the Umayyad period, certain rooms continued in use for bathing with apparently the abandonment of communal pools for small individual basins. In the eastern part of the building, a bakery for a substantial community occupied, among others, rooms S (mill) and M (large tannur). Probably at the same time, area R, which separates the thermal rooms from the street, was partly occupied by workshops.

Research on the town. Besides the excavation in the three main areas (the Nabataean Arch, the large eastern church, and the South Baths), limited observations have provided more information about the urbanization of Bosra. The remains of an east ram-


Fig. 36. Dura-Europos. The Palmyra Gate at the end of the 1992 campaign.

part have been identified. Recent town development has covered the distinct edge of the ancient city, which was still visible on early aerial photographs, but sections of wall made up of bossed masonry blocks (which are visible at the base of modern buildings) have been followed up to the burial ground that adjoins the al- Mabrak mosque. The core, and sometimes even the internal facing, of a wall at least 2 m thick was preserved. West of the mosque minaret, a return 2.8 m thick can be followed for about 20 m. Stones are bound together with clay, and the masonry is irregular, but embossed blocks were frequently found. Its date needs to be established, and unanswered questions remain: Does it belong to a late defensive system of the city, and in response to what threat was it built? And how does it relate to the earthworks around the

amphitheater? Dura-Europos. Pierre Leriche, Ecole normale su-

perieure, CNRS, Paris, and Asad Al-Mahmud, Direc- torate General of Antiquities, Damascus, report:

The 1991 and 1992 campaigns at Dura-Europos involved excavation of the Palace of the General, the

fortifications, the principal street, and in 1992, the

Temple of Zeus Megistos.33 These excavations and the related study of wall construction techniques and ceramics have yielded important information on the

history of the city and its monuments. Palace of the General. Setting up support walls to

contain the fill of the rooms in the northern facade

yielded a great deal of ceramic material important for

study of the chronology of the building. Collapse debris was removed from the slope that extends to the north of the palace, and blocks were recovered to serve in the restoration of the facade. The form of the terrain in front of the collapse was thus recovered. The work of excavation could then begin in earnest, but was limited in order not to inconvenience future excavations on the northern facade.

To the west, a sounding opened in 1990 to search for a large courtyard at the northwest corner of the

palace uncovered only the remains of an ancient ca-

nal, already excavated further to the east.

Fortifications. Work concentrated on the Palmyra Gate area (fig. 36). A sounding in the foundations of the southern tower of the building dates the gate one and a half centuries after the foundation of the city.

33 The 1991 campaign to Dura-Europos took place from 20 August to 3 October. In 1992, fieldwork was undertaken from 3 March to 2 May. The results of the research on Dura- Europos are published in the series Doura-Europos, Etudes. Vol. III is in press.

We would like to thank Y. Saint Geours, Directeur de la Division des Sciences humaines A la DGRCST; J. Leclant, Secretaire perpetuel de l'Academie des Inscriptions et Belles Lettres; G. Vallet, Secretaire de la Commission consultative des recherches archeologiques A l'tranger.


Fig. 37. Dura-Europos. The Maison sud located on the main street, level Ic.

Outside of the city, a large area of excavation was undertaken in front of the gate to clear a section of the Roman glacis.

The remains of the final battle in front of the main door of the city were uncovered here. An intense fire left a thick layer of ash and burnt debris mixed with a great quantity of objects: projectile stones, arrow and lance points, pieces of catapults, fragments of

armor, nails and pieces of war machines, and a bronze shield. The total number of metal objects recovered from a 30-m2 surface was over 400.

The chronology of the Palmyra Gate is as follows:

1) the gate was constructed in the Seleucid period (between 150 and 113 B.C.) and finished shortly before the taking of the city by the Parthians; 2) the Parthian period (113 B.C.-A.D. 165) began a long period of use without maintenance. Heavy erosion of the threshold (?) permitted the access of vehicles, but the opening of the western portal was reduced from 3.8 m to 2.8 m. A courtyard designed as a way station for people arriving at the city was built in front of the

gate; 3) during the Roman period (A.D. 165-256), a water canal coming from nearby baths crossed the

gate, and brick walls reinforced the floors that supported the war machines; 4) the Sassanians attacked the gate in 256, and the occupied city was depopu- lated.

Main street. Fieldwork has uncovered nine surfaces, the oldest of which is made of limestone and orange earth, and is identical to the first level associated with

the construction of the Palmyra Gate. This surface lies on a thick layer of organic rubbish containing ceramic material from the third century, raising the

question of the nature and function of this part of

the site during the century and a half that separates the foundation of the city and the construction of the stone enclosure and the main street.

Maison sud. The sounding of the house that borders the street to the south was continued and enlarged

(fig. 37). This house (house I) had replaced an older house (house II) probably associated with the first

surface of the street, of which only a staircase has

been found. House I opens onto the street through a door sit-

uated in its northeast corner, associated with a gypsum floor. Later, there was a fundamental change in the function of the house: the original door was walled

up and a new larger opening was cleared 3 m to the

west. A column was placed in front of the walled-up door, corresponding to a column 5 m to the south, and also most likely to two other symmetrical columns to the west of the new door; together these formed a

quadrilateral in the center of which was an altar. This

stage is associated with a series of red floors, contain-

ing ash and pits with arrowheads and lances. In the following stage (ic) the columns were re-

moved, for unknown reasons, but the function of the

building was not changed, and the red and ash floors

continued. A pilaster against the east wall appears to

have had a particular function in an area bounded by


Fig. 38. Isriye. Temple from the east, with the two rooms of the Islamic cellar in front of the entrance.

a little wall where a great quantity of ash had accumulated. On the floor, Roman coins, vases, broken

storage jars, nails, tools, and arrowheads and lances were found. This Maison sud could be seen as a Par- thian sanctuary, changed in the Roman period through removal of the columns, without which the cultural practices would have been interrupted.

Temple of Zeus Megistos. The Temple of Zeus Meg- istos, one of the oldest in Dura, was completely excavated by the Yale University Project, but never

published. In 1992, the surface was cleared again, permitting photographs and maps to be completed, and soundings were excavated to date the older levels. Some endangered structures were the object of conservation efforts.

In the course of excavation, the ceramics from the

soundings of the street rampart and in the area of the large sounding of the main street were studied. Particular attention was accorded to the earlier peri-

ods, since Hellenistic ceramics are not clearly understood.

Research on the construction methods of the enclosure wall and the citadel concentrated on a study of the vaults, arches, and lintels. A 1990 study of the blocks of houses excavated by earlier projects has led to the establishment of a systematic file of the characteristics of each block, extent, state of excavation, and the possibilities for obtaining new data.

Isriye. R. Gografe, Deutsches Archiologisches In-

stitut, Damascus, reports: The extent of information on a small temple in

Isriye has remained nearly unchanged since the short visit of H.C. Butler in 1890; however, in 1991 and 1992 the temple was cleared by a small team from the Deutsches Archaologisches Institut in Damascus, and its architecture has become completely visible.34 In the beginning of October 1992, photogrammetric documentation was undertaken.

34 H.C. Butler, Architecture and Other Arts: Publications of an American Archaeological Expedition to Syria 1899-1900 (New York 1904) 76; A. Musil, Palmyrena (New York 1928)

55-60; A. Poidebard and R. Mouterde, Le lime de Chalcis (Paris 1945) 89-94.


The results of the excavation are as follows: only the cella of the temple and its podium are preserved; virtually all of the pronaos has been carried off. Only some blocks of the foundation in the extension of the

podium are still in situ. A reconstruction of the pronaos cannot be based on the physical remains, but

only on analogous buildings. In Islamic times, a two- roomed cellar with vaulted rooms was built in place of the pronaos (fig. 38). Its roof was covered by small limestone blocks, resulting in a platform in front of the cella. The floor level in and around the cellar was the same as in the Roman period.

During the clearing of the cella, the remains of its former inner architecture came to light (fig. 39). It consisted of a row of surrounding blocks with three

projecting blocks on both long sides. In the rear we found large limestone plates, which may have confirmed the substructure of a thalamos. The above- mentioned blocks are surely the remains of a free-

standing structure with an entablature bracketing out and forming reentrant bays in between-a form dis-

tinguished from many other Roman temples in Syria, where the columns are engaged to the walls of the cella (e.g., in the small temple in Baalbek). Unfortu-

nately, we found no significant remains of such upper architecture.

In earlier times, changes took place in the cella: under the floor of the room a cellar was built, perhaps a cistern. For security reasons, we could not have a

look into this structure. The "three-aisled" cella was made into a "two-aisled" room with three bases for

supporting columns, for which we think the columns of the inner architecture of the Roman period were used. On the walls of the cella, holes for the ceiling construction of this period are clearly visible. The

building was used for living purposes until fire inter-

rupted its use. As a result of the fire, the cellar located

immediately in front of the cella was filled with ceram-

ics, bones, woven fabric, mortars, and fragments of

Arabic inscriptions of the Mameluke period, which

may date this building. Between the Mameluke period and Roman times

were other phases of use. During one of these, the

cella ceiling was apparently tiled; we found hundreds

of tiles on the north and south sides of the temple, but not in the cellar in front of it, which indicates that

there are at least two disturbance levels to distinguish. The date of the tiles is difficult to determine, but from

general comparison with Roman temple architecture in Syria, it does not seem very plausible to date them

Fig. 39. Isriye. Cella of the temple, from above.

to the Roman period. Similar tiles also occur in the

Byzantine era and later elsewhere, and we must con-

sider that the tiles may represent a phase after the

Roman period, perhaps when the building was used

as a church, even if we have no evidence other than

general analogy with other sites. Another important period is the integration of the temple into an enclosure wall that surrounded the whole settlement. This

wall, visible at the north and south sides of the cella, was the main source of the debris around the podium of the temple. It consists in its lower part of irregular limestone blocks; at its first course, the wall is approx-

imately 4 m thick. Above the limestone blocks, mud-

bricks were used. Its disturbed crown was covered by the above-mentioned tile fragments.

The problem of dating the Roman-period temple is reduced to an analysis of the architectural orna-

mentation, because the stratigraphic data are less sub-

stantial due to disturbance in later periods. This

analysis indicates a Severan date in the early third

century. We do not possess any clear information

about the deity worshipped here, since finds of cult

objects and inscriptions are lacking. Tuneinir. Michael Fuller and Neathery Fuller, St.

Louis Community College, report: The fifth season of excavations at Tell Tuneinir

took place during June and July of 1992, under the direction of M. and N. Fuller.35 Tuneinir is a 46-ha

35 Staff members during 1992 included Sharon Doerre, Dave Hanlon, Horace Hummel, Pat McWhorter, Robert McWhorter, Carl Stevens, and Orhan Nanoo. For previous reports, see M. Fuller and N. Fuller, "Tell Tuneinir of the

Habur: Preliminary Report on Three Seasons," AAS 37-38 (1987-1988) 279-90, figs. 1-7; Fuller and Fuller, "Tunei- nir," in Weiss (supra n. 12) 738-40.

158 H. WEISS, ARCHAEOLOGY IN SYRIA

site in the salvage area of the Middle Habur River

drainage. The Tuneinir project's main objective is to

investigate the range of Islamic-period buildings (houses, churches, mosques, khans, baths, etc.) for clues to the socioeconomic diversity within a 13th- 14th century town. Our second objective is to determine the size and role of Tuneinir during the Neo-

lithic, Bronze Age, Iron Age, and Classical period. In area I, we returned to deepen three excavation

units and determine the depth of a series of gray beds

deposited during the Early Bronze Age. Sterile subsoil was encountered in square 60 at an elevation of 287.3 masl. The strike and dip of the subsoil in squares 60 and 66 indicate that the first occupation on the north

edge of the site was established upon a low, natural mound of subsoil. The deepest loci in square 60 contain a combination of kitchen midden and agricultural processing waste. Important finds from loci 160017 and 160018 include two miniature clay cartwheels,

fragments of a clay cart/chariot box, small clay figurines of a goat and sheep, and a burnished clay token decorated with a series of lines and dots.

Analysis of ceramic wares from square 60 reveals that the greatest frequency of Ninevite 5 sherds occurs in locus 160013 while the greatest frequency of Me- tallic ware sherds occurs in loci 160015 and 160018. Both of these diagnostic wares are represented in the loci resting directly upon sterile soil in area I. The

majority of the potsherds from the gray beds belong to various subdivisions of coarse and cooking wares. Flint sickles, cores, waste flakes, and shatter were found in the gray beds at Tuneinir.

In area III, east of the tell, we continued excavation in the church. The three steps connecting the nave and the haikal were completely removed. Embedded within the north banister (locus 313036) were three

pieces of colored, molded plaster that match decorated fragments discovered during 1989 and 1990. The fragments discovered in 1989 were described as

pieces of a ciborium, but the new discoveries change this interpretation. We conclude that the fragments belong to an approximately 1-m long lintel that was mounted within the church above the portal between the nave and haikal. There is no evidence of weather-

ing of the paint or molded design. The central image of the lintel is an enshrined cross issuing from a pair of stylized acanthus leaves. Two pinwheel/whirl designs and two rosettes flank the enshrined cross.

Excavation beneath the floor of the haikal in locus 313033 recovered the third and final piece of a painted lintel; two pieces were found nearby in locus 313030 during 1990. The fragment discovered dur-

ing 1992 was found facing upward under the plaster paving of the top step of an internal division of the

haikal. It is decorated with a small black cross in a

geometric field of red. The original lintel was 1.5 m long; the iconography on the lintel is clearly a refer- ence to the three crosses of Calvary.

Excavation in the south end of the tell has uncovered the ruins of a khan (area IV), bath (area V), and market (area VI). Continued excavation of the bath demonstrated the presence of a hypocaust floor. The bath was supplied with water from a basalt-lined well located at the southwest of the building. The upper 6 m of fill in the well contained basalt cobbles derived from the destruction of the bath during the 12th- 13th century. The sediment beneath the destruction fill contained blue-and-black glazed sherds, cooking ware sherds, an Ayyubid-period coin, and many small rodent and bird bones.

Twelve squares were excavated in the central portion of a large ruin designated as area VI. We inter-

pret the structure as an Ayyubid-period market

measuring approximately 28 x 34 m. Pieces of iron

slag found in several rooms may reflect metalworking within area VI. Glass rods discovered in locus 658003 and a fist-sized piece of raw glass found in locus 656008 raise the possibility that a glassworker operated a small business within the market. Heavy hemi-

spherical-shaped objects of plaster discovered in several loci are evidence for small-scale yarn production using wooden swifts.

Work in the area IV khan was limited to balk removal and architectural study. The original structure

began as a square khan measuring 22.5 m on each side with interior rooms constructed in a fashion to

give mirror-image symmetry down its north-south axis. The symmetry was lost during the Ayyubid period when the west half of the northeast room was transformed into a weli. The addition of the mosque and shops, attached to the north half of the khan, also disrupted the building's symmetry. An intact, decorated stucco fragment was recovered from a col-

lapsed arch in the east half of the khan. Raised Kufic

script on the panel reads Rabb-al-SAlamin ("Lord of the Universe"). The panel discovered during 1992 is identical to two stucco panels discovered in the khan's mosque during 1990. The Arabic inscriptions from area IV and the Syriac inscriptions from area III underline the fact that Tuneinir was a multiethnic

community embracing more than one religion during the Ayyubid period.

DEPARTMENT OF NEAR EASTERN LANGUAGES

AND CIVILIZATIONS

YALE UNIVERSITY

NEW HAVEN, CONNECTICUT 06520

BITNET LEILAN@ YALEVM

archeology syria2

Documents

syrian research programs

syrian archaeology

french survey projects

french teams

collective research

aja survey

ma jor research foci

jstor archive