RESEARCH GOALS AND RESULTS
Michael D. Danti
The Tell es-Sweyhat Project has undergone many stages in terms of its research design. From 1973 to 1975, three seasons of rescue excavations were conducted by Thomas A. Holland since it was thought that the construction of the Tabqa Dam and the formation of Lake Assad would flood the site. Fortunately, this never happened. The 1970’s excavations provided critical data on the use of space in the Inner City and helped to elucidate the site’s chronological sequence (see below and Holland 2006). Excavations were renewed in 1989 and continued in 1991 as a joint expedition of the University of Pennsylvania Museum (UPM) and the Oriental Institute (OI) of the University of Chicago under the co-directorship of Richard L. Zettler (UPM) and Thomas A. Holland (OI). Following the 1991 season, this joint project was dissolved and the OI conducted a final season of fieldwork in 1992 under Holland’s direction. Penn conducted excavation seasons in 1993, 1995, 1998, 2000, 2001, 2005, and 2007 under the directorship of Zettler. This project was specifically designed to study the process of urbanization in the Early Bronze Age at Sweyhat and in the surrounding region and to investigate the site’s urban form in all its stages of development in the area of the High Mound and the Low Mound. We are still pursuing these goals, but starting in the late 1990s our emphasis began to shift toward studying agropastoralism at Sweyhat and the surrounding region, in large part as an extension of regional work started by Tony Wilkinson in the 1970s and continued in 1991–92 (see esp. Wilkinson 2004). We are no longer able to excavate on the Low Mound since a Syrian court decision in the early 1990s decided in favor of the local landowners, determining that this area was not part of the protected archaeological site, and since that time the majority of the Low Mound has been irrigated for agriculture, causing widespread destruction of the Outer City and the cemeteries of shaft-and-chamber tombs that surrounded it. In 2008, the Raqqa Museum conducted rescue excavations in many intact tombs opened by irrigation water. This project was led by Mohammad Sarhan, Director of the Raqqa Museum.
In 2008, a new project jointly sponsored by Boston University and the UPM was started under the direction of Michael D. Danti. We are still pursuing the objective of studying the site’s urban form, especially in the mid and late third millennium BC when the site underwent a radical urban transformation. In addition, we are placing added emphasis on the study of ancient subsistence practices and Mid and Late Holocene climate change in relation to regional settlement patterns and the evolution of agropastoral economies. This project also includes an archaeological field school that trains students in Near Eastern Archaeology and ancient Mesopotamia and is accredited by Boston University through the Office of International Programs.
The Ashmolean Museum Excavations of the 1970s
The 1970s rescue excavations at Tell es-Sweyhat concentrated on the Inner City and uncovered extensive remains of the late-third millennium city. The excavations revealed:
(1) Portions of the Inner City Wall, a 2.5 m thick mudbrick construction set on stone footings with buttresses or a series of defensive towers (Holland 1976: 49 and 1977: 37);
(2) Evidence for centralized grain storage (Holland 1976:55 and 59; van Zeist and Bakker-Heeres 1985: 309) and metal working (Holland 1976: 51 and 66-67), as well as administrative artifacts, including an inscribed weight (Holland 1975), in a number of interconnected rooms built against the wall on the western side of the citadel or Area IV (Holland 1976: 49-62 and 1977: 37-43); and,
(3) Domestic buildings on the northeast edge of the citadel in Area III (Holland 1976: 48-49). In addition to evidence of the late third millennium settlement, late remains were uncovered in a series of squares on the southern side of the mound, Area I (Holland 1976: 38), and stratigraphic sequences were defined in a step trench on the northern slope of the mound, Area V (Holland 1976: 62) and in two 5 by 5 m squares just west of the mound's summit, Area II (Holland 1976: 38-48).
The Penn Museum-Oriental Institute Joint Project (1989-1992)
During this second phase in the project’s history, areas first excavated in the 1970s were extended and new areas opened in both the late third millennium Inner City (most importantly Operations 1 and 5) and in the Outer City (mainly Operations 4 and 9). Work in Operation 1 in the western Inner City revealed a large structure (the Kitchen Building) that seemed to be of specialized function — it was located across a street (directly east) of a Storehouse excavated in the 1970s in Area IV. In later seasons, extending work vertically and horizontally in this part of the western High Mound would be a major objective since it held the promise of understanding the “elite” quarter of the late third millennium city — the area thought to have important secular and religious buildings at the heart of the ancient city. Work in the Outer City showed that there was minimal depth of deposit there, and that with little effort occupation phases of the late third millennium could be easily reached. The houses of the Outer City were large and contained activity areas indicative of both the household level of production and more specialized activities perhaps denoting craft specialization. For more on these seasons see especially Zettler et al. 1997. In 1992, Holland completed his fieldwork at the site. The OI’s 1992 season was primarily directed at the recovery of important wall paintings first uncovered in 1991 in Operation 5 on the southern High Mound and dating to the mid third millennium BC.
The Penn Excavations (1993-2007)
The UPM continued work for another 15 years following Zettler’s original research plan, but the project also evolved as new evidence came to light. As originally laid out in 1989, Penn's research had two interrelated major goals, one site specific and one regional. The former involved the study of form and structure in a late third millennium urban center. The regional program aimed at examining:
(1) The relatively rapid growth of the site from a small village to a major urban center in the late third millennium and its equally rapid contraction, in the early second millennium and;
(2) The degree of integration between the late third millennium urban center and its hinterland.
In approaching these research questions after the first exploratory seasons, efforts at Tell es-Sweyhat itself from 1993–98 were largely focused on the unexcavated Outer City and the regional component of the project. Yet in order to recover a range of comparably excavated Inner City contexts, it was planned in effect to cross-section the High Mound by excavating a series of 10 by 10 m units that ran west-east and south-north across it, the two seriesof squares intersecting near the mound's highest point. Completing these trenches became the project’s focus from 2000–07. On the High Mound, the project completed the east-west series of trenches with Operations 1, 12, 20, 21, 29, 40 on the west side of the mound and a small square (Operation 13) on the east side of the High Mound, and undertook smaller excavations elsewhere. The north-south trench was also completed with Operations 5, 30, 31, 34, and 41–43. The results from these trenches greatly increased our understanding of the character and development of the center of the settlement in the mid and late third millennium BC and the degree to which the urban transformation of the late third millennium BC (Sweyhat 4) represented a radical departure from the previous settlement (Sweyhat 1–3).
The Fortress Town The trenches on the High Mound have shown that in the early-to-mid third millennium BC (at least Sweyhat Periods 2 and 3) Sweyhat consisted of what we provisionally interpret as a fortress (located at the center of the High Mound) surrounded by a town estimated at 10–15 HA. The fortress had undergone many modifications and expansions over its long history and consisted of a massive mudbrick structure standing at least 5.5 m high (Danti and Zettler 2007: 177-79). The fortress is exceptionally well preserved since it was effectively buried after it was abandoned (see below). We have excavated a few of the interior chambers of the fortress and they are preserved to their roof levels. The roofs of the narrow passageways and rectangular chambers are formed with barrel arches, thrust arches, and corbelled arches. In its earliest periods of use, the fortress had projecting rectilinear towers/buttresses and the outer face of the wall was batted (sloping inward). In at least one area, the lowest parts of the wall were protected by massive limestone slabs (Operation 30). In the area of the southern High Mound, we have exposed a segment of the fortress wall measuring minimally 77 m long (east-west). The fortress is minimally 62 m north-south. We believe we now have the southeastern corner of the fortress in this span of wall, as well as a potential gateway in the southwest. There is no evidence to date that the surrounding town was protected with fortifications. Beyond the town was an area of shaft-and-chamber tombs (Zettler 1997: 51–72).
In the later third millennium BC, there was an extensive remodeling of the urban environment. This event marks the beginning of Sweyhat Period 4. The fortress was likely already falling out of use or had been abandoned. The new settlement consisted of two separated areas, the Inner City (High Mound) and the Outer City (Low Mound).
The Inner City To form the Inner City, the area of the High Mound, formerly the fortress and the structures immediately adjacent to it, was reshaped. The High Mound, measuring approximately 6 HA, was already a sharply sloping hill composed of the massive fortress and the remains of settlement at the site spanning roughly 600–700 years. The new Inner City was comprised of two levels. At the center of the settlement was the High Inner City. This area was defined by a high terrace that was formed by burying the earlier fortress and areas immediately adjacent to it. First, all the structures of Sweyhat Period 3 that had been built near and against the face of the fortress were demolished. The fortress was then buried using earthen fills and a retaining wall of stone and mudbrick was constructed to hold these fills. The area of this High Terrace stood one full building story (roughly 3 meters) above the newly leveled Low Inner City. Thus, the center of the settlement of Sweyhat Period 4 contained two tiers. We know that a long-room, bent-axis temple (the High Inner City Temple) stood at the center of the High Inner City. Other structures of ambiguous function, often poorly preserved, surrounded this structure in the High Inner City. Excavations to date indicate the Low Inner City contained storage facilities, specialized production facilities, and houses. The entire Inner City was surrounded by the Inner City Wall. This mudbrick wall, approximately 2.75 m thick, had rectangular projecting towers. In the southwest Inner City, our excavations in 2008–09 have revealed a 30m-long segment of this wall and one of its towers (see below). Structures of Period 3 were demolished to construct this wall.
The areas that have been excavated thus far show that buildings in the Low Inner City were laid out radially around the High Inner City. One circle of buildings was constructed against the face of the retaining wall of the high terrace. This inner ring of buildings was separated from a second by a street. A second ring of buildings abutted the Inner City Wall. Overall, the material culture of this part of the settlement indicates an elite presence in comparison to the Outer City and large-scale production facilities interspersed with residential areas.
The Outer City An Outer City surrounded the two-tiered Inner City (see topographic map). The project made controlled surface collections; conducted a remote sensing mapping project; and, undertook excavations in eight areas of the Low Mound (Operations 3-4, 9, 15-18, 23, 25, 27, and 37). The Outer City covered approximately 24 HA according to our most recent estimates. The area contained large houses and production facilities such as pottery kilns (Operations 16 and 23). The settlement was surrounded by the Outer City Fortifications — chiefly an earthen rampart measuring 18.5 m wide (excavated in Operation 25). The rampart was probably surmounted by a fortification wall, although no traces have been found of it due to erosion. The rampart’s outer face had a sloping stone revetment. Beyond the rampart was a ditch. In the northwestern Outer City we have found different types of fortifications — a mudbrick casemate wall (Operations 15 and 18). Perhaps the Outer City had multiple phases in its fortification(s) or simply incorporated several forms of fortification contemporaneously. The project also carried out limited excavations (Operations 19, 26, and 28) in the Southern Extramural Settlement immediately beyond the Outer City Fortifications. In this area, estimated at 4-5 HA, poorly preserved traces of features and structures were recovered.
Parts of the Outer City were first occupied in Sweyhat Period 3; however, it seems that there was a sudden expansion of the settlement in Sweyhat Period 4, at which time this part of the settlement became a distinct, circumvallated area. The Outer City was occupied into the late third/early second millennia BC (Sweyhat Period 5). During the final Period of Sweyhat’s Bronze Age occupation (Sweyhat Period 6) the Outer City was steadily abandoned as the settlement contracted to the area of the High Mound. During Period 6, the settlement’s fortifications fell out of use and the site was a small town/village. The last two periods of Sweyhat’s occupation (Sweyhat Periods 5 and 6) remain poorly known since they are not well preserved in most areas excavated to date.
Regional Research Since its inception, the Tell es-Sweyhat Project has had a regional component (see below). One of the most interesting aspects of Tell es-Sweyhat is its location — 4km from the Euphrates in an arid area (250 mm of rainfall per year with 25–35% interannual variability) where canal irrigation is impracticable. Rainfall agriculture is very risky here, yet it formed an important part of the Early Bronze Age economy. Moreover, the site did not occupy an important position on a known ancient trade route or control precious natural resources. Yet, Sweyhat thrived in the mid and late Early Bronze Age, especially, oddly enough, during a period of heightened aridity starting around 2200 BC. We often refer to this situation as the Sweyhat Enigma. Sweyhat was not alone in its seemingly precarious location; other large sites in neighboring valleys were similarly located. During this period, settlement expanded beyond the fertile, well-watered banks of the Euphrates River onto to neighboring main terrace (where Sweyhat is located) and even into the undulating area of limestone hills, a forbidding environment indeed, between the Euphrates and the Balikh. This unusual settlement pattern was only equaled in the Late Roman to early Islamic Period and in the modern era. We call this settlement pattern the Sweyhat Paradigm, and we believe it is indicative of increased emphasis on pastoral production within an agropastoral economy coupled with strong local government authority.
In the 1970s, Tony Wilkinson began an intensive geomorphological and archaeological survey aimed at reconstructing long-term land use patterns in the Sweyhat embayment and completed his work in 1991 and 1992 (see esp. Wilkinson 2004). In 1993, Michael Danti initiated a follow-up regional research program which involved expanding Wilkinson's areal coverage and making topographic maps, systematic surface collections, and soundings at relevant third and early second millennium sites (see Regional Surveys and Excavations). This project was primarily aimed at reconstructing the ancient pastoral economy since Wilkinson’s work had been focused on reconstructing agriculture (primarily dry farming). This project showed that there was abundant evidence of ancient settlement in the arid upland region between the Euphrates River and the Balikh River to the east, including Early Bronze Age occupation of the late third millennium. The main phases of occupation in this upland zone mirrored those in the Sweyhat embayment. Danti attributes the presence of sites in this region, not an important area for agriculture, to pastoral activities (Danti In press). One of the main achievements of the regional project was the excavation of Tell Hajji Ibrahim. This small site (0.25 HA), located approximately one kilometer from Sweyhat, was excavated in 1993, 1995, and 1998 under the direction of Danti (2000). Tell Hajji Ibrahim was found to date from 3100–2800 BC and through most of its history served as a fortified grain storage facility, probably for a largely transhumant pastoralist population. The site was later used as a cemetery in the mid-third millennium BC (Sweyhat Period 3) and in the Seleucid and Late Roman periods. Tell Hajji Ibrahim is one of the few sites excavated in northern Syria that has evidence for the earliest Bronze Age, and in fact occupation at the site overlaps the Late Uruk presence at nearby sites such as Jebel Aruda. Levels dating to the earliest Bronze Age are present at Tell es-Sweyhat, but they are situated beneath thick layers of later occupation and are thus difficult and costly to reach. Hajji Ibrahim and other small sites like it were abandoned at Sweyhat started to grow in late Period 2 and Period 3, which is interpreted as a process of sociopolitical and economic centralization centered on Sweyhat.
In summary, the 1993–2007 excavations provided a general overview of Sweyhat’s urban development during the Early Bronze Age, particularly the latest periods of the Early Bronze Age (Sweyhat Periods 4–6). Sweyhat was founded in the late fourth to early third millennia BC, and was probably composed of a central fortress surrounded by a small village/town for much of its early history. This settlement slowly grew in the early part of the third millennium and the pace of growth quickened in the mid third millennium as settlement on the main terrace was centralized at Sweyhat. Then, in the mid-to-late third millennium BC (early Sweyhat Period 4) the site was transformed into an urban complex with a two-tiered, fortified Inner City and a fortified Outer City. Interestingly, this transformation and Sweyhat’s peak period of settlement correspond to an episode of climatic deterioration in the Near East (see esp. Weiss et al. 1993; Weiss 2000; Staubwasser and Weiss 2006). During Sweyhat’s urban florescence, there is evidence for a widespread destruction(s) — the High Inner City Temple and storehouse of Area IV were burned and buildings in other areas of the site were burned or intentionally razed to the ground with their contents largely in situ (see below). We are as yet unsure whether this was a single event and are still refining the chronology. In most areas, the destruction was followed by an episode of rebuilding along the original architectural plans. In the latest third and early second millennia, Sweyhat began to decline in size, and eventually the site was abandoned. This period of decline corresponds to a regional shift in settlement patterns to smaller scale settlements located on and near the floodplain of the Euphrates River that typifies Middle Euphrates settlement patterns in the Middle and Late Bronze Age. The area of the High Mound was again occupied in the Seleucid and Late Roman periods.
The Boston University-University of Pennsylvania Museum Excavations (2008–present)
Since 2008, the Tell es-Sweyhat project has shifted its research emphasis to investigating heretofore unexplored areas of the Low Inner City with several related objectives in mind:
1) To refine our understanding of the Period 4 destruction phase or phases that were first documented in Area IV and the High Inner City Temple;
2) To investigate the cultural response(s) to the putative 4.2 ka aridification event documented in global proxy climate data;
3) To better understand the Low Inner City in terms of the use of space and the sociopolitical and economic units associated with it;
4) Refine our chronology of Sweyhat Periods 3–6;
5) Better understand the critical transition from the Sweyhat Period 3 settlement to that of Sweyhat Period 4;
6) Better document the Seleucid and Late Roman period settlements.
To do this we are currently conducting excavations in three areas: the southwestern High Mound, the southern High Mound, and the eastern High Mound.
The Southwestern High Mound (Operations 101–104, 150, and 151) In 2008, we opened new operations on the southwestern slope of the High Mound in an area we rightly predicted would be the southwestern corner of the Low Inner City. Thus far we have completed two seasons of excavation in this area along the western face of the Inner City Wall. The sequence of Early Bronze Age occupation in this area is similar to that recovered in the 1970s in Area IV, consisting of, from lowest levels reached to the surface,
1) Poorly preserved Sweyhat 3 architecture — the remains of structures that were demolished to build the Inner City Wall;
2) Well preserved structures of Sweyhat Period 4 abutting the earliest phase of the Inner City Wall that were partially burned and razed to the ground with large numbers of objects left in situ;
3) A partial rebuild of the earlier structures (Sweyhat Period 5), usually directly atop the wall stubs of the Sweyhat 4 and abutting the Inner City Wall;
4) An ephemeral layer of EB-MB transitional occupational debris (Sweyhat 6) in the areas of higher elevation;
5) Stone lined cist graves and a “bathtub burial” of the Seleucid and Late Roman Periods.
In 2009, we sectioned the city wall in two locations to study the construction techniques employed, its history of use, and its dating. Contrary to interpretations advanced by Holland (2006), the city wall in this area was in use in both Sweyhat Periods 4 and 5. The wall was not abandoned until Period 6. In the area excavated to date, we have a 60 m span of the wall cleared and the base of one tower partially exposed. Within the structures built against the inner face of the city wall, we have identified a bakery, residential areas, and a small temple or household shrine. We are collecting large numbers of paleoenthnobotanical samples and soil samples for micromorphological analysis to further investigate the activities that were connected with these spaces and climate during Sweyhat Periods 4 and 5. We also have large numbers of radiocarbon samples from this area and are awaiting results of the dating of the Bronze Age levels. In 2010, we hope to continue to dig deeper in this area to reveal more of the Sweyhat Period 2–3 settlement.
The Southern High Mound (Operations 100 and 200) Excavations and observations of surface remains in the 1970s indicated that the southern High Mound had well preserved remains of the Seleucid and Late Roman Periods. Test excavations in the 1970s and in the early 1990s verified this. The Seleucid–Late Roman represents a peak period of settlement in the area on par with the Early Bronze Age. Since investigating the cultural responses to climate change and the subsistence economies that supported urban settlement in this arid region is a major focus of our research, detailed examination of these periods provides valuable comparative data for the Early Bronze Age dataset. In 2008 and 2009, we began the clearance of large parts of several Late Roman structures in this area. The ceramics and other artifactual finds are quite different from those recovered at the summit of the mound, where the uppermost level appears to date to the Seleucid period. The Roman occupation on the southern High Mound is substantial and seems to be a well laid out village of the early centuries AD. Results from 2009 indicate that this settlement extended to the eastern High Mound.
The Eastern High Mound (Operations 110a–c) In 2009, we began excavations on a prominence on the edge of the eastern High Mound. This area had not been previously explored. Our objectives in this new area are to locate the line of the Inner City Wall and gather information on the use of space in this part of the Low Inner City to compare with our work on the southwestern High Mound. The stratigraphic sequence in this area is similar to that in our other trenches:
1) A Late Roman structure of the early centuries AD;
2) An ephemeral layer of EB-MB transitional, including large ash pits;
3) Well preserved storage structures of Sweyhat Period 4 built against the retaining wall of a terrace running east-west. A stairway led up to the top of this terrace. The storage structures were burned and the contents, primarily shell beads, were strewn about the surrounding area;
4) Earlier Bronze Age remains in small soundings of as yet undetermined date.
In 2010, we plan to expand our horizontal exposure in this area to recover more of the well preserved Late Roman structures and the Sweyhat Period 4 buildings in this area, particularly structures that might have been built atop the terrace to the north.