Tell es-Sweyhat is a large 3rd millennium site on the Euphrates River in northern Syria. Excavations undertaken in the mid-1970s (Holland 1976, 1977) suggested the site as a case study in urban growth and collapse and a nearly ideal “field laboratog” for studying early northern Mesopotamian cities. In 1989 the University of Pennsylvania Museum requested an excavation permit and conducted a short field season at the site. Penn and the University of Chicago’s Oriental Institute then farmed a joint project, with cooperative excavations in the fall of 1991 and alternate year field seasons thereafter. This report and the accompanying contributions from project team members focus on Penn’s work at the site and in its vicinity. A recent survey in The Economist (July 29, 1995) noted that cities are booming and argued that their growth is economically healthy and culturally beneficial. Regardless of editorial perspective, the figures are staggering. Nearly half of the world’s population and three-quarters of Westerners live in cities. In the last thirty years the number of city-dwellers worldwide rose by 1.4 billion and over the next fifteen years it will rise by around one billion more. Most of the growth occurred in developing countries, but during the past decade, cities in industrialized countries—even old downtowns—have enjoyed a resurgence. Such figures suggest the process of urbanization and urbanism as compelling, if not critical, research areas and, as tong-lived and generalized phenomena, areas that demand historical and cross-cultural perspectives. As the “cradle” and arguably the most urban of the world’s primary civilizations, Mesopotamia merits particular attention. Indeed, southern Mesopotamian, with its irrigation agriculture base, has been dubbed “heartland of cities” (Adams 1981). The process of urbanization in southern Mesopotamia, which climaxed in the late 4th millennium B.C., has been the focus of intensive archaeological investigation (Adams 1981), and major projects analyzing ancient urban forms in the area are currently underway (Stone and Zimansky 1995). By way of contrast, the dry farming regions of northern Syria and upper Mesopotamia were, for a long time, thought of as a backwater. In the last twenty years, however, excavations have shown that complex urban and literate civilizations existed there too in the 3rd millennium. Although the process of urbanization in the area has been the subject of scholarly speculation (Weiss 1990; Weiss et al. 1993; Weiss et al., n.d.), the topography of northern Mesopotamian cities remains poorly understood. The Tell es-Sweyhat Project was undertaken to explore both the shape of the city and the process of urbanization in these northern settlements.
Tell Es-Sweyhat
Tell es-Sweyhat lies on the east bank of the Euphrates, 65 kilometers down river from Jerablus (Figs. 1, 2). It sits in the center of an embayment, a crescent-shaped plain surrounded on the north, east, and south by a plateau of desert steppe and bordered on the west by the slightly entrenched flood plain of the river (Fig. 3). The embaynient, a mid-to-late Pleistocene river terrace, takes in 48 square kilometers of arable land. It lies within the semiarid dry farming zone; however, dry farming is precarious, so pastoralism, probably in the past as today, would have been a critical part of the area’s subsistence economy (see Miller’s and Weber’s reports). The site (Figs. 1, 4) consists of a central mound surrounded by an outer, or lower, town and enclosed by a rampart or wall. The site’s core area is 35 hectares (87.5 acres), but the ancient settlement may have been somewhat larger. Aerial photographs show a dark line that would appear to be a wall or rampart abutting the south side of the outer fortification and enclosing an area 250 by 450 meters. This southern extension would increase the size of the settlement by more than 10 hectares (25 acres) (see below). The 1970s excavations focused on the dentral mound, with a stratigraphic step trench down its north slope (Area V) and a deep sounding to its west (Area II). Late or early 2nd millennium houses on the northeast were exposed (Area III), as well as portions of an inner fortification wall, apparently curving with the mound’s 10.5 meter contour, and a large building constructed against the wall on the western side of the mound (Area IV). The building revealed evidence for centralized storage of foodstuffs, e.g., barley and field pea (van Zeist and Bakker-Heeres 1985), and for metal working. A long-term goal of the current series of excavations is understanding the process of the settlement’s seemingly rapid growth from village to urban center. With work at the site focused largely on the late 3rd millennium occupation, our approach to the process of the settlement’s urbanization was planned to be regional in scope (see Danti’s report). Nevertheless, both by design and accident, excavations have provided insights on the ancient settlement’s development over time.
Sequence of Occupations
An outline of Tell es-Sweyhat’s growth and development had emerged by the end of these excavations. We could see that the site was occupied by the beginning of the 3rd millennium. Although initially a relatively small village covering the area of the central mound, it expanded to a 35-hectare settlement in the final centuries of the 3rd millennium. ‘The original village became a fortified citadel; a substantial outer town emerged around it, and the urban area was surrounded by an enclosure. By 1993, excavations in Operation I (Figs. 5), on the west side of the main mound, had reached the late 3rd millennium levels that lay more than 2 meters below the mound’s surface. In beginning a new phase of regional work (see Danri’s report), we were hampered by the lack of a detailed ceramic sequence into which we could tie other sites in the embayment. We took advantage of the situation by clearing the western end of the operation to sterile soil; in 1995 we broadened the sounding. This deep sounding, along with one in Area II, has produced ceramic evidence that relates Tell es-Sweyhat to other sites in the embayment and helps establish a sequence of phases within the site itself. We have divided Operation 1’s sequence into eight occupation phases. The earliest occupation, Phase 1, yielded pottery similar to that from the lowest phases in the Area II sounding and dates to the early 3rd millennium. Phase 4 belongs to the late 3rd millennium; Phases 5-8 postdate the late 3rd millennium occupation, but probably end before 1800 B.C. since they yielded no Middle Bronze IIB ceramics. Our sequence has shown that the early 3rd millennium village probably extended to the western edge of the main mound. However, occupation there is considerably lower than the apparently contemporary occupation in the Area II sounding. The disparity in elevations may suggest that the core of the mound conceals an even earlier occupation. If not, it may be that the early 3rd millennium levels in the center of the mound were raised, perhaps by setting buildings on artificial platforms or terraces. At the other end of the timeline, the presence of Phases 5-8 suggests that the settlement did not completely collapse in the late 3rd or early 2nd millennium. Occupation continued for a time, with population concentrated on the central mound. Excavations in Operation 25, a long, narrow trench in the eastern outer town across the outer fortification wall, revealed that the wall was built on the leveled remains of earlier houses that date to the third quarter of the 3rd millennium. The houses suggest that the mid to late 3rd millennium settlement might have covered as much as 10 hectares. The settlement’s growth, then, was more continuous than had been assumed.
A Mid 3rd Millennium Cemetery
A serendipitous discovery revealed the existence of a cemetery on what would have been the edge of the mid to late 3rd millennium settlement. In 1993, the owner of the land on which Tell es-Sweyhat sits began irrigating in the area of the northwestern outer town. The water flow opened up several deep holes in the ground, revealing the location of tombs. We began salvage excavations in two of the open holes: Tomb 1 and Tomb 2. Tomb l’s rectangular shaft, cut from sterile soil, was about 2 meters deep (Fig. 6). It ended on a ledge, from which steps led down to the floor of a chamber about 5 meters in diameter and 2.5 meters high. Tomb 2 had partially collapsed, and we could not determine its size or exact shape. Both tombs had been looted. Tomb l’s shaft, however, was sealed by the outer fortification wall. This wall, built at the end of the 3rd millennium, provided a terminus antequem for the tombs, while the pottery and small finds suggested a date in the third quarter of the 3rd millennium. In the following season, we noticed a sink hole in the tomb field. Taking this to mark the location of an unlooted tomb, we began excavations in the area. Our perception paid off. Beneath the surface lay Tomb 5, a multiple burial of at least ten individuals, accompanied by hundreds of artifacts—personal ornaments, weapons, animal hones, pottery, and more (Figs. 7, 8; see also box on Tomb 5). Shaft-and-chamber tombs with multiple burials are common in the Euphrates River valley (Orthniann 1980). The Sweyhat tombs are spread over an area of at least one hectare. Based on their size the cemetery may include 100 to 150 tombs. Whether it was associated with the settlement or served as burial place for a regional population is uncertain Nevertheless, such a major unlooted cemetery is a potentially rich source of information on the mid to late 3rd millennium population and demographics, as well as on diet and disease. If the Biblical account of Jacob’s instructions to his sons at his death (Gen. 49: 29-33) can be taken as documenting similar burial practices known from the Levant, the tombs probably represent family burials. Analyses of skeletal remains focusing on features assumed to have a genetic basis, supplemented by genetic data that might he provided by the analysis of ancient DNA, might shed light on the relationship of the individuals in the tombs (Brown and Brown 1994; Thuesen 1995). Artifact distributions might provide insights into the socio-economic structure of the early settlement.
Urban Form and Structure
Sweyhat is the reconstruction of the topography of the late 3rd millennium urban center and the elucidation of the political, socio-economic, and ideational rules that it encoded. The goal was suggested in part by A. Leo Oppenheim’s classic discussion of differences between southern and northern Mesopotamian cities. Southern cities, with their diffuse power structures, had similarly diffuse layouts in which temples and palaces were spread across the city. Northern cities consisted of a fortified center containing the palace and temple with their dependencies, and an outer residential area commonly surrounded by a second fortification wall. Oppenheim dubbed them “citadel cities” (1964:130-33). Cuneiform tablets from the mid 3rd Gillennumb royal palace at Tell Mardikh (ancient Ebla) provide evidence seemingly consistent with Oppenheim’s description (Matthiae and Petrinato 1976; Archi 1990). Archaeological work, notably at Tell Gaya and Tell.
Data Recovery in the Outer Town
Our work in the outer town involves both noninvasive data recovery techniques and excavation. We were particularly interested in whether we could identify specialized activity areas and features with these methods. We began by making systematic surface collections in the outer town and area south of the outer rampart. In 1993, we undertook a pilot program to determine the feasibility of geomagnetics for creating useful archaeological maps of the outer town. Our pilot program was highly successful (see Peregrine’s report) and we continued the geomagnetic mapping in 1995. We have now mapped 7 hectares or 17.5 percent of the outer town and outer town south (see Fig. 4), including large areas on the east (Block 1), northwest (Block 8), and south (Block 9). The geomagnetic data has added to our understanding of the topography of the outer town. Block I (see Peregrine, Fig. 3), for example, shows the outer and inner fortification walls, as well as a gateway in each wall; a network of streets; room blocks, including an industrial area of kilns; and what may he a large open space to the north of the room blocks. Our excavations include several test exposures undertaken to determine the “ground-truth” of the geomagnetic anomalies revealed in the mapping project. Perhaps channeled water from a nearby well. We found some evidence for the existence of wells in the form of Euphrates river cobbles on the surface to the south of the operation. The source of a water supply has been one of the more perplexing aspects of Sweyhat’s location; it lies 3 kilometers from the Euphrates river, the nearest natural source. Perhaps a number of wells would have existed in the settlement much as they do in the villages of the area today. The inhabitants would have obtained drinking water directly from wells and stored it in jars in their houses, as the pebble-lined pits in the Operation 4 building suggest. The conduits would have provided water for other purposes and permitted waste water to be flushed away. Our geomagnetic survey initially picked up marked anomalies in the eastern outer town near a surface collection unit in which we found a kiln waster. Operation 16 (see Peregrine Figs. 1, 2) located the sources of three markedly high magnetic anomalies that were clustered closely together and revealed parts of six rooms or courtyards. The most prominent features found were the fireboxes of horseshoe-shaped kilns; wasters and fused sherds suggested the kilns were used for firing pottery. The Operation 16 kilns and walls showed up clearly in the geomagnetic maps of the area. Operation 23, immediately southeast of Operation 16, revealed a well-preserved circular kiln, circa 5 meters in diameter, in a shallow depression in sterile soil. We also tested areas where our geomagnetic survey picked up no noticeable anomalies. Operation 17 to the north, for example, yielded only a series of pits cut from a surface that had probably been eroded or plowed away. The area may have been an open trash dump in the late 3rd millennium. In any case, the excavation supports our supposition that parts of Block I represent undeveloped space, substantial areas of which, if we can generalize, would appear to have existed in the lower town. We had uncovered part of the settlement’s outer fortification in clearing Tombs 1 and 2 in 1993. We undertook more work on the wall in 1995 to test the results of our earlier mapping. Specifically, we tested parallel linear anomalies circa 20 meters apart in Block 1, which we posited represented the line of the eastern outer fortification wall. Operation 25 exposed several occupation phases. The second phase consisted of the outer fortification wall, which, as already noted, was built on the leveled-off remains of the earlier buildings. In contrast to the northern outer fortification wall, apparently a casemate structure, the eastern wall was a 20-meterwide earthen rampart, faced on the outside with a sloping stone revetment and on the inside with a mud-brick wall set on stone footings.
Data Recovery South of the Outer Fortification Wall
Sherd densities in the area south of the outer rampart were lower than densities in the outer town, but Operation 19B, located over stones visible on the surface and mapped in our survey, yielded the fragmentary stone footing of a 1.5-meter-wide wall. The archaeological deposit in the area was less than 50 centimeters. Our work yielded few diagnostic pottery sherds. In general, survey and geomagnetic data, as well as excavations, have combined to prove that the area south of the outer wall was occupied at the time of the site’s flourishing.
Conclusions
A major difference between applied and “pure” research is the open-endedness of the latter. Archaeological projects with broad research goals such as the Tell es-Sweyhat Project require an indeterminate number of field seasons. As work proceeds, digging problems come and go, some general questions get answered and unanticipated possibilities arise. We have now begun to answer some of the questions we posed when we initiated the project, but we have nevertheless just “scratched the surface.” The major immediate challenge ahead involves continuing long-planned work at the site, as well as in the region, while responding to the demanding, hut unparalleled opportunity of studying the mid 3rd millennium settlement’s population which a newly discovered Tell es-Sweyhat cemetery presents.
Acknowledgements
The 1993 field season at Tell es-Sweyhat vas funded by the National Endowment for the Humanities and the 1995 field season by the National Endowment for the Humanities and the National Geographic Society. The authors wish to thank James Amstrong of the Harvard Semitic Museum for his contributions to the project.