(originally published in: (1996) XVII Kazi Sonuclar Toplantisi. Pp. 109-128. Ankara: Republic of Turkey, Ministry of Culture, General Directorate of Monuments and Museums.)
1994 EXCAVATIONS AT HACINEBI TEPE
Gil J. Stein and Adnan Misir
INTRODUCTION:
The third field season of the Joint Sanlõurfa Museum-Northwestern University salvage excavations at Hacinebi Tepe, Birecik district, Sanlõurfa province took place from July 6-August 27, 1994, co-directed by Gil Stein (Northwestern University) and Adnan Misir (Sanlõurfa Museum). The excavations were funded with support from the National Endowment for the Humanities (grant number RK-20133-94), the National Geographic Society (grant number 5295-94), the Kress Foundation, Northwestern University, Dr. Joseph and Laura Kiser, and other private donors.
We wish to express our appreciation to Prof. Dr. Engin Özgen, General Director of the Ministry of Culture’s Directorate of Monuments and Museums for permission to conduct this research. We thank Hamza Güllüce, who served as representative of the Sanlõurfa Museum. We also gratefully acknowledge the assistance of Kemal Isik, the Muhtar of Ugurcuk village for his hospitality and assistance in conducting excavations at Hacinebi.
In addition to the two co-directors, the project staff consisted of: Shahira Ashkar (University of Arizona), Tonia Callas (conservation intern), Cheryl Coursey (State University of New York-Binghamton), Cynthia Cripps (conservation intern), Dr. Christopher Edens (Harvard University), Barbara Foster (Archaeological Illustrator), Fokke Gerritsen (University of Chicago), Sarah Hinds (Cambridge Archaeological Unit), Alan Lupton (Cambridge University), Nancy Mahoney (Arizona State University), Dr. Augusta McMahon (University of Chicago), Dr. Naomi Miller (University of Pennsylvania), Jeffrey Nicola (Northwestern University), Julia Pearce (University of Pennsylvania), Dr. Holly Pittman (University of Pennsylvania), Margaret Reid (Guilford Technical Community College), and Katherine Untch (Asian Art Museum).
SITE DESCRIPTION AND 1994 PROJECT GOALS:
Hacinebi Tepe is a 3.3 ha. low mounded site located on the bluffs overlooking the east bank of the Euphrates river in Sanlõurfa province, southeast Turkey (figure 1). The site lies near the head of the main north-south riverine trade route linking Mesopotamia and Anatolia; it also occupies a strategic location on what has historically been the major east-west river crossing point at Birecik. The mound of Hacinebi is situated on an easily defensible east-west oriented spur which drops down steeply to the river river on west, and into deep canyons to north and south.
Three seasons of excavation have documented three main occupations Hacinebi Tepe (Stein 1994). 1-2.5 m. thick Persian/Hellenistic deposits (ca. 5th-2nd centuries BC) are present immediately below the plow zone. Beneath these are Late Chalcolithic deposits up to 5 m. deep, dating approximately 3900-3200 BC. The Late Chalcolithic occupation consists of two phases: an earlier Phase A - ca. 3900-3700 BC which has only Local Anatolian Late Chalcolithic ceramics (the “Pre-Contact” phase), and a later phase B ca, 3700-3200 BC (the “Contact Phase”), which has both Mesopotamian Uruk style ceramics and Local Anatolian Late Chalcolithic ceramics. The Hacinebi excavations investigate the effects of contact with Mesopotamia on local cultures of southeast Anatolia in the fourth millennium BC. The 1994 field season had six main goals:
1) Broad horizontal exposures of the Possible Uruk Area in the Contact Phase B Occupation;
2) Broad horizontal exposures of the Local Anatolian Public Architecture in the Contact Phase B;
3) Broad horizontal exposures of the Local Anatolian Occupation on the West Slope;
4) Analysis of the Late Chalcolithic Seal Impressions;
5) Analysis of Ceramics and Chipped Stone Tools;
6) Conservation of metal artifacts found in the 1993 season.
In the 1994 season, six trenches were excavated, exposing an area of approximately 600 m2 (figure 2). Excavations took place in three different parts of the mound: the north area (operations 4, 6, and 10); the south area (operations 7 and 11); and the west area (operation 5). These units were intended to remove the overlying Hellenistic deposits and to excavate the two main Late Chalcolithic occupation phases at Hacinebi.
ACHAEMENID/HELLENISTIC PHASE:
The Achaemenid/Hellenistic occupation at Hacinebi dates from the fifth to the second century BC, and is the latest major settlement at the site. Deposits from this period were recovered in all three excavation areas.
In the west area, excavations in operation 5 exposed a series of large trash pits filled with Hellenistic ceramics, animal bone and other artifacts, but no in situ architecture from this period. In the north area, Achaemenid/Hellenistic deposits were exposed in operation 10. A modern concrete and stone cistern in the center of this 10 x 10 m trench was removed in order to expose the underlying archaeological deposits. The cistern cut through a large Hellenistic mud brick building which originally extended north into the immediately adjacent operation 4. The walls are constructed of square mud bricks on a stone foundation course, and are oriented northeast-southwest. In some places the walls are preserved to a height of approximately 1.5 m. The monumental size of the structure and its proximity to the remains of a defensive wall found in operation 6 in the 1993 field season suggests that this was a public building of some sort, possibly part of a military outpost. Based on the find of a copper coin of Alexander the Great on the floor of this structure in operation 4, we can date the entire complex to the late fourth-early third centuries BC. The large structure was later cut through by a series of deep, straight sided, mud-plaster lined pits, probably dating to the late third and second centuries BC. The careful construction and fairly regular spacing of these pits suggest that they may have been used for grain storage. At some point in the late second century, these pits were re-used - some for burials, and others for trash disposal.
Achaemenid/Hellenistic deposits were also recovered from operation 11 in the south area of the mound. Operation 11 was a 10 x 10 m trench placed along the northern edge of operations 2 and 7. Beneath the plow zone, excavation revealed a 1.2 m. deep Hellenistic occupation layer, consisting mainly of a massive planned architectural complex which is almost certainly a public building (figure 3). The mud brick walls are 1.7 m. wide and are oriented northeast-southwest. These walls are constructed on a stone foundation set into a trench. The northeast mudbrick wall of the room is slightly out of alignment with the stone foundation course; this apparently reflects the need to compensate for the potential instability of the wall due to its construction on the south slope of the mound. The Hellenistic building was remodelled at least twice. In the latest of these re-modellings, four large pithoi were set into the floor of the main room exposed in this structure. The deposits inside the building complex contained very few artifacts, suggesting that the building was cleaned out and abandoned. The large scale Hellenistic architecture is cut through by a series of pits, apparently dating to the 3rd-2nd centuries BC. These pits are the latest occupation in Operation 11.
LATE CHALCOLITHIC PHASE B (CONTACT PHASE):
Underlying the Achaemenid/Hellenistic occupation at Hacinebi are deposits dating to the Late Chalcolithic phase B, which corresponds to Amuq F (Braidwood and Braidwood 1960), Arslantepe VIA (Frangipane and Palmieri 1989; Palmieri 1985, Frangipane 1993), and Kurban VIA (Algaze 1990) in the middle and late fourth millennium BC. We have called Hacinebi phase B the “Contact phase” because it is characterized by the presence of Mesopotamian Uruk style ceramics and other artifacts alongside, but often separate from the local southeast Anatolian artifact assemblage. The high concentration of Uruk ceramics, administrative artifacts, clay wall cones, and other characteristic Uruk artifacts such as cruciform grooved stone weights, bitumen, and clay sickles in the north area of the site suggests that a small group of Mesopotamians may have lived as an enclave of foreign merchants at Hacinebi during the contact phase. Phase B deposits have been found in all areas of the mound excavated to date.
In the north area of the mound, the 1994 excavations recovered Late Chalcolithic phase B remains in Operations 4, 6, and 10.
Earlier excavations in Operation 4 had exposed residential architecture dating to the middle and end of the Contact phase B. The 1994 excavations explored the underlying earlier Contact phase B building levels in this area. The ceramics from these buildings consist mainly of Local Anatolian Late Chalcolithic types, although Uruk Mesopotamian style Beveled Rim Bowls are also present.
Excavations in operations 4 and 10 exposed the northern half of large free standing stone platform over 2.4 meters high (figure 4). This upper platform lies immediately to the west of the Lower Terrace in Operations 1 and 6. The Upper Platform is rectangular, over 9 m long from NE-SW, and 8 m from NW - SE. The limestone-filled filled platform appears to have been used for a long time, and underwent a series of repairs and remodellings. As a result, the original architecture on top of the stone foundations is no longer preserved. However, the monumental size and height of the Upper Platform suggest that the original superstructure was either a public building such as a temple, or perhaps an elite residence.
Excavations showed that the upper platform was constructed first, and that the lower terrace was added at a later date to broaden the area at the top of the mound. A small ramp and stairway connect the two structures. In its earliest phases of use, the upper platform was free standing, surrounded by a large open area.The earliest Contact phase B ash deposits (loci 179/181) lying on the broad outdoor surface to the north of the upper platform yielded a large number of local Anatolian style stamp seals and exclusively Anatolian ceramics, with no traces of Mesopotamian Uruk pottery. The southern side of the upper platform showed signs of a hiatus in occupation before later repairs and remodelling took place. In one of the major remodellings, wall 89/94 was constructed, forming a narrow alley along the south edge of the platform. In the later portion of the Contact phase B, a series of stone and mud brick domestic rooms were built up against the north face of the upper platform, and eventually on top of it as well. A large pit (locus 159/180) cut into these later building levels yielded 47 local Anatolian style stamp seal impressions associated with exclusively local ceramics. These seal impressions resemble the seal impressions found in the 1993 season in mostly depicting cervids and lions (figure 5). These motifs are common on southeast Anatolian stamp seals at sites such as Arslantepe (Ferioli and Fiandra 1983; Frangipane and Palmieri 1989) and De©irmentepe (Esin 1990). In addition to the seal impressions, a carved local style limestone stamp seal was found in operation 10. The seal depicts two animals separated by a curvilinear design. Two unfinished limestone blanks for hemispheric stamp seals indicate that local style stamp seals were actually manufactured at Hacinebi.
Some of the most important finds of the 1994 season were recovered from pit 54, which dates to the late portion of the Contact phase B occupation of Operation 10. Pit 54 contained exclusively Uruk Mesopotamian ceramics in association with two cylinder seal-impressed clay jar stoppers and a cylinder seal impressed clay tablet (figure 6). Only half of the tablet was preserved, so it is impossible to say whether or not it originally contained numerical notations. This is the only known example of an Uruk period tablet from Turkey. A blank, unused clay tablet was also found in the same pit. The jar stoppers, tablets, and cylinder seal impressions, combined with the 1993 discovery of the bulla and tokens, show that the full range of Uruk Mesopotamian administrative artifacts was in use at Hacinebi. The Uruk tablet and jar stoppers were found in association with the one local Anatolian style stamp seal impression on a clay container closure. This is important because the presence of both Mesopotamian and Anatolian sealings in the same pit provides clear evidence for direct exchange between the two groups at Hacinebi.
In the south area of the site, phase B deposits were exposed in operations 7 and 11. Operation 7 is a 10 x 10 m. trench on the south slope of the mound, immediately to the west of Operation 2. The trench was first excavated in 1993. In 1994, excavations completed the exposure of Contact phase B deposits down to the latest building levels of the Precontact phase A. Excavations focused on the area to the south and west of stone platform 26. Most of operation 7 appears to have been an open area at the south edge of the mound during the Contact phase B. These excavations led to the redating of stone platform 26 and the niched mud brick building (walls 44,48,55) first discovered in 1993. These structures were initially thought to date to the Contact phase B. However, additional excavations and analysis of the associated ceramics showed that both were in fact constructed late in the Precontact phase A. Stone platform 26 remained in use during the Contact phase B, providing strong evidence for continuity between the two Late Chalcolithic occupations at Hacinebi.
In operation 11, the earliest phase B building level reached in 1994 was a monumental structure with well built stone walls over 1 m. thick (figure 7). The most notable feature of this building is a series of stone 2 m. wide niches and buttresses on the east (inner) face of the wall. This massive structure was clearly a public building - either a temple or an administrative structure. Later in the contact phase B, the structure was subdivided into series of small, probably residential rooms with narrow, poorly constructed stone and mud brick walls. In the latest part of the phase B occupation, a series of pits was excavated in operation 11, including a long, narrow pit which was later filled with Uruk beveled rim bowls. This evidence suggests that the function of this part of the site changed markedly during the later part of the contact phase B. Overall, it appears that the total inhabited area of Hacinebi decreased in size over the course of the contact phase B.
LATE CHALCOLITHIC PHASE A (PRE-CONTACT PHASE):
The earliest occupation known so far at Hacinebi is Late Chalcolithic phase A, dating to the early fourth millennium BC. Hacinebi phase A is chronologically equivalent to Amuq F (Braidwood and Braidwood 1960) and Arslantepe VII (Frangipane 1993) Phase A deposits directly overlie sterile soil in all areas of the site excavated to date. Phase A is also called the “pre-contact phase” because it is defined by a purely local ceramic and artifact assemblage, with no evidence for trade or any other kind of interaction with Mesopotamia. In the 1994 field season, phase A deposits were exposed in the north and west areas of the mound.
Operation 6 is an 8 x 10.5 m. trench in the north area of the site, immediately to the south of Operation 1. This trench was first excavated in 1993 exposing large areas of a 2.6 m. high platform terrace (the “Lower Terrace”) with large amounts of Mesopotamian Uruk style artifacts. The construction of this terrace platform in Operations 1 and 6 marks the beginning of the Contact phase B at Hacinebi. The 1994 fieldwork completed the excavation of Late Chalcolithic deposits in the east end of operation 6 down to sterile soil. Excavations in Operation 6 had two main results. First, they confirmed that Uruk Mesopotamian ceramics are associated with the earliest use of the Lower Terrace at the beginning of Contact phase B. In addition the Operation 6 excavations exposed more of the large Precontact phase A wall at the eastern edge of the mound. The wall is 1.5 meters wide and 1.2 meters high The wall is oriented Northwest-Southeast, and is constructed on a deliberately laid bed of compact gravels and silts as a levelling foundation. The top of the wall is flat and probably served as the footing for a thick mud brick wall which is no longer preserved.Test cuts behind wall 110 showed that it was built as a terrace set into the sterile soils of the small natural hill on which the initial Late Chalcolithic settlement of Hacinebi was founded. The purpose of wall 110 remains unclear. It may have been used for either defense or as a levelling terrace for the buildings of Precontact phase A.
Pre-contact phase A deposits were also exposed in the west area of the site, in operation 5. The 1994 excavations expanded operation 5 from a 5 x 5 m to a 10 x 10 m exposure, and recovered a well stratified sequence of phase A building levels. The ceramics from the upper building levels of Operation 5 are almost entirely Local Late Chalcolithic chaff tempered wares. Uruk Mesopotamian ceramics and cylinder seal impressed jar sealings were present only in pits 75 and 31, which apparently were dug late in the Contact phase B.
The lower building levels can be dated by their associated ceramics to the Precontact phase A. Expansion of Operation 5 exposed two large, narrow rooms constructed of stone walls 1 m. thick (figure 8). These structures appear to be storage rooms of some sort, adjacent to the mud brick courtyard house exposed in the 1993 excavations. In one room of the mud brick house, excavations recovered a ceramic mold with traces of copper, and a copper chisel (figure 9). These finds provide clear evidence for metallurgical activity and long-distance procurement of copper ores at Hacinebi during the Precontact phase.
Finally, the 1994 excavations in operation 5 recovered a phase A local style stamp seal impression from a context which was radiocarbon dated to between 4100 and 3770 BC. The reverse of the clay sealing shows that it was used to seal the cover of a woven basket. This is our first evidence for local administrative activities in the pre-contact phase at Hacinebi.
RADIOCARBON DATES AND CHRONOLOGY:
Twelve radiocarbon samples provide information about the dating of the Late Chalcolithic occupations at Hacinebi (figure 10). All samples were processed by Beta Analytic Laboratories in the United States, and were calibrated using the CALIB 3.0.3 radiocarbon calibration program (Stuiver and Reimer 1993). Four of these dates derive from the pre-contact phase A. Two of the phase A samples (Beta-67912 and Beta-67914) give late fourth and third millennium dates that appear to be far too recent. However, the remaining two samples (Beta-67913 and Beta-80944) both give consistent dates in the early fourth millennium BC.
The eight samples from the contact phase B permit us to date the period of Uruk Mesopotamian presence at Hacinebi. Two of the samples (Beta-80942 and Beta-80940) derive from pits which contained a tablet, two mushroom-shaped jar stoppers, and container closures, all bearing Uruk Mesopotamian cylinder seal impressions. This glyptic material can now be securely dated between 3700 and 3500 BC. A pit filled with beveled rim bowls also yielded a date in the same time range (Beta-80943). This is important because it indicates that the Mesopotamian presence at Hacinebi begins in the Middle Uruk period. Other Uruk sites in Syria such as Tell Sheikh Hassan (Boese 1987) and Tell Brak (Oates 1986; Oates and Oates 1993) also date to the Middle Uruk period, centuries earlier than the Late Uruk dates for colonies such as Habuba Kabira, Jebel Aruda, and Hassek Hüyük (Behm-Blancke 1986, 1992, Sürenhagen 1986, Van Driel and Van Driel-Murray 1983). These radiocarbon dates indicate that the period of Mesopotamian contact with Syria and southeast Anatolia began earlier and lasted longer than we had previously thought. The remaining five radiocarbon dates from Hacinebi suggest that the site was occupied at least up to 3300 BC and possibly later.
LABORATORY ANALYSES:
In addition to the field excavations, several important artifact studies were also conducted in the 1994 field season. Cheryl Coursey examined the ceramics from Operations 4, and 7 in order to refine the typology and chronology of Local Late Chalcolithic ceramics at Hacinebi. Dr. Holly Pittman recorded and analyzed the seals and more than 100 sealings from the 1993 and 1994 seasons. Dr. Christopher Edens studied over 5000 chipped stone artifacts from Operations 1, 2, 5, 6, 7 and 10 in order to determine whether different types of economic activities took place in the northeast, south, and west areas of the site. Dr. Naomi Miller processed soil samples to remove carbonized plant and wood remains as part of her study of prehistoric agricultural systems and land-use patterns at Hacinebi. Katherine Untch, Tonia Callas, and Cynthia Cripps stabilized and cleaned the metal artifacts from the 1993 excavations. Cleaning of bronze artifacts from tomb 28 in Operation 7 showed that at least one signet ring was Achaemenid in style, permitting us to assign a tentative 5th century BC date to this burial. Cleaning of a coin from the 1993 excavations in Operation 4 showed it to be a bronze single inscribed with the name and image of Alexander the Great, thereby dating the large Hellenistic building complex of Operations 4 and 10 to the late 4th century BC.
CONCLUSIONS:
The 1994 field season at Hacinebi exposed large portions of the Contact phase B Late Chalcolithic settlement dating ca. 3900-3200 BC, spanning both the Middle and Late Uruk periods. The finds of Uruk style wall cones, stone weights, ceramics, and cylinder seal impressed administrative artifacts such as a clay tablet, bulla, tokens, jar sealings and jar stoppers provide additional evidence for the presence of Mesopotamians living in the local Anatolian settlement at Hacinebi. The occurrence of both Uruk and local Anatolian seal impressions in the same pit provides clear evidence for exchange between the two groups. It is hoped that further excavations in 1995 will be able to clarify the nature and extent of the Pre-Contact phase A at Hacinebi.
REFERENCES CITED:
Algaze, Guillermo, ed.
1990 Town and Country in Southeastern Anatolia, Volume 2: The Stratigraphic Sequence at Kurban Höyük. Chicago: University of Chicago, Oriental Institute Publication 110.
Behm-Blancke, Manfred
1992 Hassek Höyük Eine Uruk Station im Grenzland zu Anatolien. Nürnberger Blätter zur Archäologie 8:82-94.
1986 Die Ausgrabungen auf dem Hassek Höyük im Jahre 1985. In VIII Kazõ Sonuçlari Toplantõsõ. Pp. 139-147. Ankara: T.C. Kültür ve Turizm Bakanlõ©õ Eski Eserler ve Müzeler Genel Müdürlü©ü.
Boese, J.
1987 Excavations at Tell Sheikh Hassan, Preliminary Report on the Year 1987 Campaign in the Euphrates Valley. Annales Archéologiques Arabes Syriennes 36/37:67-100.
Braidwood, Robert, and Linda Braidwood
1960 Excavations in the Plain of Antioch. Chicago: University of Chicago Oriental Institute.
Esin, Ufuk
1990 De©irmentepe (Malatya) Kalkolitik Obeyd Evresi Damga Mühür ve Mühür Baskõlarõ. In X. Türk Tarih Kongresi. Pp. 47-56, plates 33-41. Ankara: Türk Tarih Kurumu Basimevi.
Ferioli, Piera, and Enrica Fiandra
1983 Clay sealings From Arslantepe VI A: Administration and Bureaucracy. Origini 12(2):455-509.
Frangipane , Marcella
1993 Local Components in the Development of Centralized Societies in Syro-Anatolian Regions. In Between the Rivers and Over the Mountains. M. Frangipane , H. Hauptmann , M. Liverani , P. Matthiae , and M. Mellink , eds. Pp. 133-161. Rome: Universita di Roma "La Sapienza".
Frangipane, Marcella, and Alba Palmieri
1989 Aspects of Centralization in the Late Uruk Period in Mesopotamian Periphery. Origini 14:539- 560.
Oates, David, and Joan Oates
1993 Excavations at Tell Brak 1992-93. Iraq 55:155-199.
Oates, Joan
1986 Tell Brak: The Uruk/Early Dynastic Sequence. In Gamdat Nasr. Period or Regional Style? U. Finkbeiner and W. Röllig, eds. Pp. 245-273. Wiesbaden: Dr. Ludwig Reichert Verlag.
Palmieri, Alba
1985 Eastern Anatolia and Early Mesopotamian Urbanization: Remarks on Changing Relations. In Studi di Paletnologia in Onore di Salvatore M. Puglisi. M. Liverani, A. Palmieri, and R. Peroni, eds. Pp. 191-213. Rome: Universita di Roma: La Sapienza.
Stein , Gil, and Adnan Misir
1994 Mesopotamian-Anatolian Interaction at Hacinebi, Turkey: Preliminary Report on the 1992 Excavations. Anatolica 20:145-189.
Stuiver, M., and P.J. Reimer
1993 University of Washington Quaternary Isotope Lab Radiocarbon Calibration Program Rev. 3.0.3A. Radiocarbon 35:215-230.
Sürenhagen, Dietrich
1986 The Dry-Farming Belt: The Uruk Period and Subsequent Developments. In The Origins of Cities in Dry Farming Syria and Mesopotamia in the Third Millennium BC. H. Weiss, ed. Pp. 7-43. Guilford (CT): Four Quarters Publishing Co.
Van Driel, G., and C. Van Driel-Murray
1983 Jebel Aruda, the 1982 Season of Excavation. Interim Report. Akkadica 33:1-26.
Figures
Figure 1: Map of NE showing location of Hacinebi.
Figure 2: Site map showing 1994 excavation areas.
Figure 3: Operation 11- Achaemenid/Hellenistic mud brick and stone public building.
Figure 4: Operations 1, 4, 6, and 10: Late Chalcolithic upper platform.
Figure 5: Phase B Local style carved limestone stamp seal and most common seal motifs.
Figure 6: Operation 10: Contact phase B Uruk tablet and jar stoppers
Figure 7: Operation 11: Late Chalcolithic monumental stone building with niches and buttresses.
Figure 8: Operation 5: Late chalcolithic phase A storerooms and houses.
Figure 9: Operation 5: Late chalcolithic phase A ceramic mold with traces of copper, and a copper chisel.
Figure 10: Hacinebi Tepe Late Chalcolithic Radiocarbon dates.
Gil J. Stein
g-stein@northwestern.edu
Anthropology Department, Northwestern University
Last modified - July 20, 2001