Visit Us | Contact Us | Membership | Make a Gift | Calendar | Order Online | What's New

Print this Page

Home > Research > Publications > News & Notes



By K. Aslihan Yener, Assistant Professor of Archaeology, and
Tony J. Wilkinson, Research Associate (Associate Professor)
The Oriental Institute
The University of Chicago

(This article originally appeared in The Oriental Institute News and Notes, No. 148, Winter 1996, and is made available electronically with the permission of the editor.)

It was with great anticipation that teams from the Oriental Institute returned to the Amuq Valley in the state of Hatay, Turkey, after a 57 year hiatus. Between 1932 and 1938 the path breaking Amuq Valley Surface Survey by Professor Robert Braidwood and his Chicago coworkers found 178 mounds. The subsequent excavations, under the direction of Calvin Wells McEwan, at Hatay-Çatal Höyük, Tell al-Judaidah, Tell Tayinat, Tell Dhahab, and Tell Kurdu provided a stratigraphic sequence upon which most of the chronology of the Near East depends even today. In 1991 Professor Douglas Esse of the Oriental Institute initiated a project to reexcavate Çatal Höyük through a collaboration with the Antakya Museum. Unfortunately his untimely death delayed undertaking the excavation. Our renewed Amuq efforts began in 1995 and were intended to be the first phase of a multi-phase regional investigation.

The 1936 field team in front of Rehaniah (Reyhanli) House, Amuq Headquarters, (left to right, standing) Robert Braidwood, Hamilton Darby, Calvin McEwan, (seated) Arthur Pierson, Elizabeth McEwan, Rigmor Jacobsen, Thorkild Jacobsen, Harold Hill, and Richard Haines.
Our enthusiasm for reactivating the Amuq projects was matched only by the excitement and support of the local population, who greeted us with warmth and hospitality. The city of Reyhanl, where Tell al-Judaidah and Dhahab are located, extended honorary citizenship to the Braidwoods and sent them a plaque thanking them for their research contributions in their region.

The 1995 field season on the Amuq Plain was divided into two operations: The first, directed by K. Aslihan Yener with graduate students Elizabeth Friedman and Clemens Reichel, was a rescue excavation that consisted of section cleaning, section drawing, and limited excavations at Tell al-Judaidah, which would enhance the stratigraphy from previous excavations and explore the possibility of future wide scale investigation. The second, directed by Tony Wilkinson with graduate students Scott Branting and Jerry Lyon, was an ongoing program of archaeological and geomorphological surveys in the Amuq Plain. They located the mounds and other archaeological sites that were mapped by the original Chicago team in 1932, after which they attempted to reconstruct landscape changes through time and to assess the impact of human communities on the environment (see below).

The 1995 field team featuring the Mayor of Reyhanli, (left to right, bottom row) Mehmet Hazirlar, Elizabeth Friedman, Aslihan Yener, Mayor Mahmut Ekmen, Bekir Altan, (top row) Clemens Reichel, Scott Branting, Tony Wilkinson, and Jerry Lyon.

Excavations at Tell al-Judaidah (K. A. Yener)

The Amuq Valley has held great interest to me ever since our ongoing lead isotope analysis program indicated a source of metal for several bronze and silver artifacts excavated previously from the Amuq. Several samples of copper and silver ore from mines in the nearby Taurus and Amanus Mountains, located to the west of the Amuq, matched Chalcolithic, Early Bronze, and Late Bronze Age metal artifacts, indicating that they were made from these ores. Of special interest were the Amuq G figurines from Tell al-Judaidah, which are to date the oldest tin bronzes in the Near East (ca. 3000 B.C.). Finding the metal workshops where the actual crafting was done lent immediacy and relevance to this project.
Robert Braidwood excavating a burial at Tell al-Judaidah, 1936.

Excavations at Tell al-Judaidah, 1995.

Bulldozer cuts at Tell al-Judaidah had exposed substantial mudbrick architecture on its northern and northeastern edges. Ceramics in the section and sherds collected from the debris indicated that the walls dated to roughly Phase F (earlier fourth millennium B.C.) and Phase G (later fourth millennium and early third millennium B.C.). Three architectural levels were clearly apparent and were recorded in a 1:20 scale drawing. The earliest mudbrick walls had been highly vitrified due to burning, which resulted in good preservation, while the later phase walls were less coherent. A sounding 5 meters in width and 2 meters deep was initiated on top of the bulldozer cut. Thus a stratigraphic column would hopefully correlate with the established ceramic sequence as well as help generate an independent Amuq F-G transition sequence, enhanced by dates from radiocarbon samples.
Team members Elizabeth Friedman and Clemens Reichel leveling in archaeological strata at Tell al-Judaidah.
Massive mudbrick walls with stone foundations (brick size: 52 x 36 x 7 centimeters) were preserved to a height of 1.80 meters and 1.50 meters in width in the uppermost phase. A room of 2.60 x 2.50 meters formed part of a substantial architectural unit. Several occupation surfaces emerged with predominantly wheel-made Plain Simple Ware dating this later structure to Amuq G. After the removal of the foundations, the sounding descended in a narrower vertical column measuring 2.90 x 1.40 meters, which was positioned just above an earlier burnt room visible in the section. The next phase contained ashy lenses, a stone alignment, and another mudbrick wall with stone foundations oriented northeast-southwest along the line of the section, but very little of this was recoverable.
The southern end of Tell al-Judaidah showing the 1995 excavated area.

Multiple-brush painted ware from the Phase G levels at Tell al-Judaidah.

The earliest floor in the building phase visible in the cross-section was not reached; however, a later mud floor with gypsum patches was recovered below the collapse level. A large ashy lens overlying several burnt destruction layers containing disintegrated mudbrick collapse represented the lowest architectural phase. Three substantial walls 1.60 meters wide and 1.50 meters high bordered a room that may have been used as a magazine or storage room and was perhaps part of an administrative unit in Phase G. Large quantities of crushed pottery were recovered on the floor of the storage room. New types of storage jars and cooking pot wares distinguished it from the phases above. Hitherto unknown Plain Simple Ware storage jars with a cream-buff paste were decorated with a red wash on the interior of the rim with red paint drizzling down vertically on the outer surface of the vessel. This decoration is not attested in the published pottery assemblage of Phase G and may belong to an earlier level of Phase G or part of an undetected F-G transition. Although several cooking pot wares and chaff-faced examples in this room do suggest a Phase F date, the more well-known Phase F strata will probably emerge from below the surface level of the present plain.

In future years the archaeological heritage of the area will be examined within the context of important technological, subsistence, and cultural changes taking place in the ancient Near East and will provide the basis for our understanding of the cultural history in this unique environment bounded by resource-rich mountain highlands.

Lake Antioch Unplugged (T. J. Wilkinson)

Our main aim during the survey part of the 1995 field season was to obtain an overview of the region from the estuary of the Orontes River inland to the plain itself in order to formulate research goals for future seasons and to get an impression of how the area had changed since the 1930s when Professor Robert Braidwood conducted his groundbreaking survey.

Although our survey was hardly groundbreaking, other people had been breaking ground, hence we witnessed the poignant sight of several tells literally cut in half or otherwise mutilated as local farmers and landowners enlarged fields to grow more irrigated cotton. Earth-moving machines, although threatening to the existence of archaeological sites, can also provide a bonus for field archaeologists by supplying opportunistic sections through deep accumulations of soil. Such sections enable us to infer environmental changes from the layers of sediments that have accumulated through time under varying hydrological, climatic, or vegetated conditions. One of the key environmental changes that would have influenced the growth of settlement in the region must have been the development of Lake Antioch, which occupied the central part of the basin in recent centuries.

Scott Branting dwarfed by the mass of Site 31 - cut in half by bulldozing operations.
As noted by Professor Braidwood, Sir Leonard Woolley, and others, this lake may have been a relatively late feature, but if it was late, we needed to ask what the basin was like before the lake formed, and how the inhabitants may have lived under such different conditions. Although we get some hints about the lake's history from Islamic writers and even from Assyrian reliefs, it seemed logical that we also examine the landscape itself for traces of its early history. By taking advantage of the fact that the lake has now been drained, the 1995 field season was able to provide some clear evidence that related to such questions. Particularly significant was the discovery of archaeological sites within the main body of the former lake. Because such sites have their occupation levels well below the water level of Lake Antioch, it follows that when they were occupied the lake must have been much smaller.

The two newly-discovered sites were recorded together with what must have been a buried ancient land surface. The sites occur some 1.5 to 2.0 kilometers to the north of the lake margin as it existed in the 1930s (see map of major sites) at a point where they must have been covered by some 2 meters of water. Hence it was possible for Woolley to declare that from a boat it was possible to lean over the gunwale and peer through the clear waters to see the ruins of Hellenistic, Roman, and Byzantine houses and churches. Such ruins evidently must have pre-dated the lake's formation. One such site--perhaps the site witnessed by Woolley--was visited by the Oriental Institute team during the field season. It was discovered by asking the local inhabitants whether they knew of any archaeological sites specifically within the area of the former lake. The main newly-discovered site "Tell al-Hijar" was, it turned out, quite well known and we first heard reference to it while working well to the north of the lake in the vicinity of Qara Tepe (Site 86). We then went southward, asking for such a site en route until we eventually tracked it down, to the north of the village of Tell ed-Diss located on the former lake margin. The site when discovered was seen to occupy some two hectares rising to an impressive altitude of about 1 meter above the lake floor. Only an ant could seriously describe the site as a tell, and the only significant traces of occupation (apart from the abundant pottery) were large stones that appear to have been eroded out from probable Middle Bronze Age buildings. The range of occupation was quite impressive, however, ranging from at least the early part of the Early Bronze Age (i.e., early third millennium B.C.) through Middle Bronze Age, Hellenistic, Roman, Byzantine, and the final occupation being approximately of early Islamic date. This range of occupation was extended back to the Late Chalcolithic and perhaps even Neolithic by the nearby site of AS 181 (nobody had the audacity to call this scrawny patch of sherd-infested cotton field a tell). Although during the later periods of occupation it is possible that Tell al-Hijar might have formed an island within the lake, it is unlikely that the lake level can have been as high as it was in the 1930s when Braidwood surveyed the area; we would therefore argue that the mid-sixth century date for the inception of the lake suggested by Woolley is incorrect, and rather it did not attain its maximum size until after the eighth or ninth centuries A.D. This would call into question Woolley's historical dating, which blamed the formation of the lake on the 551 A.D. earthquake at Antioch. However, the lake was certainly a major feature in the writings of both Yakut and Abu-l Fida, who wrote in the thirteenth and fourteenth centuries A.D.

Map of major sites within the southern Amuq Plain and the submerged site of Tell al-Hijar (adapted from OIP 48, maps 6-8); map of general area inserted.

Tell Dhahab from Tell al-Judaidah. this small site on the edge of Reyhanli had approximately one third of its bulk removed by earth-moving activities.

Soil sections exposed in the base of the former lake indicate that during long periods of time the valley axis must have been a moist environment with waterlogged, probably marshy soils. Therefore we can infer that when Tell al-Hijar was occupied, there was no major lake, but instead there existed a mosaic of marsh interspersed with localized expanses of open water. By the Middle Bronze Age such a wet valley floor environment must have extended as far east as Qara Tepe (Site 86), which Professor Braidwood described as being "in the deepest part of the marsh" (as opposed to within the lake itself). At this site, sections exposed by recent bulldozing showed that walls of the mid-second millennium B.C. were made of two types of mudbrick, both of which must have been excavated from soils in the immediate vicinity. First, a red-brown oxidized soil had been dug from dry and rather freely drained soils, and secondly, gray clay containing small freshwater snail shells was excavated from a much more waterlogged and marshy soil. The latter building material clearly indicates that wet and marshy conditions must have existed in the area during the Middle Bronze Age. Hence from this wall alone we can infer that the site of Qara Tepe was surrounded by a mosaic of drier freely-drained soils appropriate for cereals and other crops as well as waterlogged marshy and/or permanently wet soils that may indicate areas where reeds were procured and aquatic resources, e.g., fish, may have been obtained. Future work by the team will attempt to describe such environments more accurately.

A significant outcome of this environmental reconstruction is that a number of major archaeological sites, which in the recent past were surrounded by water or marsh, would have been within dry or partially dry land during the third and second millennium B.C.

The floor of Lake Antioch some forty years after draining. This is actually the site of Tell al-Hijar which may be seen as a few stones within the cotton.
Whether the lake was the result of a major earthquake dislodging a vast mass of rock to block the Orontes River, as Woolley contended, or was a more gradual process is uncertain at present. However, that a gradualist view seems more appropriate comes from the ninth century B.C. gates of Shalmaneser III at Balawat in northern Iraq. These gate decorations of bronze executed in repoussé technique illustrate a number of fortified settlements within areas of lake or marsh in the land of Unqi. Thus if the Amuq Plain is indeed Assyrian Unqi, then by the time of Shalmaneser's campaign it is evident that the area was swampy and must have formed a shallow lake in places.

Such settlements are illustrated by L. W. King in the Bronze Reliefs from the Gates of Shalmaneser (1915) and are described by A. T. Olmstead in History of Assyria (1923), who states:

For the collection of this tribute, it was necessary to penetrate the great swamp of Unqi, access to which could only be gained by flat-bottomed boats that could pass anywhere in the shallows. Shalmaneser did not trust himself to their uncertain protection, but contented himself with a position on the shore across from where, on a low mound in the midst of the swamp, stood the capital, a double-gated fortress with battlements on its walls (pp. 127-28).

Because lakes can be formed as a result of a variety of circumstances, it is premature to postulate a single mechanism behind the formation of Lake Antioch. Lakes can develop as a result of changes in climate (increased rainfall or decreased evaporation), changes in runoff from the catchments that supply the water, or from changes in the river channels themselves, as well as from a number of other human or natural causes. An additional mechanism is the process of riverine deposition, because it is evident that the Orontes River has built up a considerable sedimentary accumulation in the form of a river levee along its course, with the result that the coarse sediments have aggraded closer to the river and finer sediments have accumulated farther away. By the time one reaches the valley axis one is therefore in a true flood basin that is several meters below the level of the levee upon which the river flows. During floods, therefore, water overspills the banks and accumulates in the flood basin to the north which then becomes the feature known as Lake Antioch.

Middle Bronze Age wall that contains freshwater mollusks indicative of a second millennium B.C. wet environment around Qara Tepe (Site 86).

Thick accumulations of riverine silt and clay mask the landscape near Tell Atchana. Fourth millennium B.C. levels are buried at 3.5-4.0 meters depth near water level.

Just as Woolley describes some 5.0 meters of sedimentation at Tell Atchana (Alalakh), recently excavated drainage channels provide sections which demonstrate that alongside Atchana some 3.5 meters of alluvium/levee has aggraded since probably the late Chalcolithic period (that is since 3000 B.C. or so). Thus the Orontes levee effectively forms a barrier across the southwestern end of the Amuq Plain, a basin, which also contains the earlier phases of the Afrin River as well as the Kara Su, both of which must have exited the basin farther to the north along the present Kara Su drainage canal (the Kara Su drains from the north, the Afrin from the east, and the Orontes from the south). Together with the rising and aggrading Orontes, these rivers must have caused the plain to rise thereby causing flood waters to accumulate and marshy and eventually open water conditions to increase through time. Although tectonic uplift (perhaps associated with earth movements such as the 551 A.D. event) may have encouraged lake formation, present evidence suggests that sedimentary aggradation by restricting drainage may have been the primary cause of lake formation. In addition, the construction of irrigation canals directly from the Orontes to the valley floor must have resulted in more water flow into the valley floor. This, combined with massive overbank floods, which are a constant problem along the Orontes, might easily account for the formation of the lake.
A chisel mold of soft serpentine stone from Tell Rasm located neat the Orontes River. Whether this really belongs to the Neolithic levels of the site as would be suggested from its context or is actually later remains a subject for future investigations.
From our brief visit to the Amuq Plain, we can therefore see that it has been a marshy--and locally open--water environment for much of the last few thousand years. However, the lake was certainly smaller during earlier phases of occupation during the third millennium B.C. Therefore, sites that Professor Braidwood had to wade through water to reach in the 1930s must have been partly surrounded by dry land when they were occupied. Other sites, such as Tell al-Hijar, which were totally submerged and inaccessible, even by boat, during the original survey must have been within wet valley floors, but how wet this was is a subject for future investigations. Furthermore, the considerable amount of sedimentation on the plain may have buried other sites beneath alluvial silt and clay. Such sediments are, however, not totally bad news for the archaeologist, because by containing pollen and other environmental data, these can in turn throw valuable light upon environmental change, a subject that we hope to amplify in future field seasons.
K. Aslihan Yener is Assistant Professor in the Oriental Institute and the Department of Near Eastern Languages and Civilizations at the University of Chicago. Her article on Tin Smelting (co-written with Bryan Earl) appeared in News & Notes #146 (Summer 1995).

After training in Canada in high Arctic geomorphology, T. J. Wilkinson became interested in the archaeology of the Near East. He worked as a free-lance archaeological consultant for several years and became Assistant Director of the British Archaeological Expedition to Iraq, Baghdad. He joined the Oriental Institute as a Research Associate in 1992.

John Larson, Oriental Institute Museum Archivist, provided the two pictures from the 1930s excavations at the Amuq.

Revised: July 30, 2007

Home > Research > Publications > News & Notes