Contextual Analysis of the Use of Space at Two Near Eastern Bronze Age Sites

Part 2: Sampling an Urban Centre: an Introduction to Tell Brak Excavations 1994-1996

Roger Matthews

Contents

Location of the Site

History of the Project

The Archaeological Strata and Methodology

The Excavations

Early-Middle Uruk Period

Early Ninevite 5 Period

Ninevite 5 Temple

Other Ninevite 5 Levels

Later 3^rd Millennium / Akkadian Period

Post-Akkadian Period

Bibliography

List of contexts

Location of the Site

History of the Project

Tell Brak was first excavated by Sir Max Mallowan in the 1930s (Mallowan 1947). Among the many discoveries made by Mallowan at Brak were a temple of late 4^th-early 3^rd millennia date - the so-called Eye Temple - an enormous palace of Naram-Sin, of late 3^rd millennium date, and other architecture and finds ranging in date from the 4^th to the 2^nd millennia. In the mid-1970s the challenge of Tell Brak was taken up by Professor David Oates who conducted many seasons of excavations, focussing on the massive public buildings of the later 3^rd millennium and on unravelling the stratigraphy of the 4^th and 3^rd millennia. A palace of 2^nd millennium date on the mound's summit was also revealed (Oates et al. 1997).

In sum, previous work at Tell Brak had demonstrated the importance of the site during, in particular, the late 4^th-early 3^rd millennia and the later 3^rd millennium. It was also very clear that the site constituted an especially sensitive marker of cultural interactions between northern and southern Mesopotamia, with periods of intense contacts separated by episodes of more localised development. The main aim of the 1994-1996 programme of investigations at Tell Brak was to develop a long-term view of the shifting cultural interactions attested at the site.

The Archaeological Strata and Methodology

The first step in the methodology was the collection of surface objects over selected areas of the mound. By far the commonest object on the mound's surface is pottery. Based on the quantities of sherds recovered from surface operations it is estimated that the surface of Tell Brak is currently covered in anything between 5 and 10 million sherds. The recovery of surface material was designed to give information concerning the distribution across the site of the various periods. In all the areas where we conducted surface collection followed by scraping and excavation we found high correlations, in terms of dating, between surface pottery and sub-surface architecture.

A second field technique was that of surface scraping, whereby the uppermost crust of earth was removed using wide-bladed hoes and the new surface cleaned and recorded in detail. On the heavily eroded spurs this technique on the whole worked excellently and we were able to plan a large number of intact features directly below the surface of the mound. Visible features included walls, floor sequences, brick-packing, pits, ovens and spreads of rubbish. Again we were able successfully to use this information to direct the location of excavation trenches.

The bulk of information comes from excavation, as presented below. An important part of our field strategy has been the conduct of an experimental project under the remit of the Leverhulme Trust, entitled "Contextual analysis of the use of space in two Near Eastern Bronze Age sites". This three year project is concerned to characterise, in both quantitative and qualitative senses, context types as encountered in the excavations at Tell Brak and in the Bronze and Iron Age levels of Kilise Tepe in southern Turkey. Samples from context types, such as room fills, courtyard deposits, occupation deposits on floors, pit fills, street and rubbish deposits, are processed and analysed using directly compatible techniques so that the complex processes of settlement and society, such as the distribution of activities within domestic architecture, the extent and use of public spaces, the degree and nature of interaction between the urban and rural sectors of society, may all be thoroughly investigated.

The Excavations

Initial indications were that the large spur at the north-west of the mound, designated Area HS, held an intact and readily accessible sequence of occupation spanning at least the earlier part of the 4^th millennium until the very late 3^rd millennium, and we focused much of our efforts on this spur. Over the three years we have explored a total of six areas of the spur.

Early-Middle Uruk Period: At the base of the spur levels of Early-Middle Uruk date were investigated, that is early to mid-4^th millennium. In the lowermost trench, HS6, a large mud-brick wall, with multiple rebuilds, may have functioned as an imposing boundary to a compound or large-scale structure of some sort. The continual rebuilding and repairing of this wall may be some indication of its importance through time. A series of pottery kilns, with pot wasters, show a degree of industrial activity on what was probably the fringes of the settlement. The pottery from this trench is distinctive and has its best parallels in the Early to Middle Uruk pottery of Tepe Gawra level XI in northern Iraq. Here, then, we already have evidence, in the form of a substantial compound wall and pottery kilns, for a large-scale, supra-domestic element in the settlement at Tell Brak in the early 4^th millennium.

In trench HS1, adjacent to HS6, a series of structures of Middle Uruk date were excavated. Rarely for Tell Brak some human burials, all of children, were encountered, and pottery typical of the Middle Uruk period was recovered in great quantities. A major group of finds was retrieved from an ashy layer in a lane, comprising a number of clay sealings with seal impressions. On their obverse faces they bear impressions made by cylinder seals. These small pieces of clay have amongst the earliest securely stratified cylinder seal impressions from anywhere in the Near East and are thus of considerable significance. Cylinder seals were thought to have originated in south Mesopotamia and to be intimately connected with the development of proto-cuneiform writing and urban complexity in the later part of the 4^th millennium. And yet here we have unmistakable evidence for the use of cylinder seals in north Mesopotamia by the middle of the 4^th millennium. Given the size of the Middle Uruk settlement, or group of settlements, at Tell Brak and the increasing evidence there for social and administrative complexity, we may continue to query the traditional view of northern Mesopotamia as a late benefactor of south Mesopotamian-inspired cultural influence.

Early Ninevite 5 Period: Moving on to Tell Brak in post-Uruk times, the Ninevite 5 period is a key episode in the late pre-history of northern Mesopotamia and we have little understanding of its true nature. The collapse of the Late Uruk world system, which reached from south Mesopotamia into Syria and beyond, was followed by a renewed period of regionalism across all of Mesopotamia. At Brak we have excavated several structures of Ninevite 5 date, ranging from its beginning to its end.

Dating to the start of the Ninevite 5 period, around 3,000 B.C., we exposed in trench HS4 an impressive structure with a series of plastered floors, the uppermost of which was burnt. After its uppermost floor had been burnt, the building had been carefully packed with mud bricks of two colours. Several pits had been dug through the brick packing but stopped at floor level. The function of this building is not known, but the quality of the floors and the treatment of the structure after its abandonment do not suggest purely domestic activities.

Other early Ninevite 5 evidence was revealed in trench HS2 in the form of a series of ash-filled pits, yielding quantities of early Ninevite 5 pottery, including a rare example of a painted chalice. Also from the pits came a collection of clay sealings with cylinder seal impressions. The function of these sealings is largely concerned with the closure of baskets, showing a concern to control movement of portable goods within these containers. The style of the seal impressions is typical of Ninevite 5 sealings known from elsewhere, especially Nineveh itself.

Ninevite 5 Temple: Later Ninevite 5 occupation was investigated further up the spur in the upper levels of trench HS4. Overlying the brick-packed building mentioned above, a series of stratified surfaces and rubbish deposits had accumulated, from which came many examples of the fine incised pottery and coarse wares which typify the later Ninevite 5 period. These deposits had built up through long-term use of an open courtyard which was associated with a series of single-roomed structures to its east. This complex constitutes one of the most important Ninevite 5 buildings yet uncovered in Mesopotamia.

We excavated two phases of the enclosed complex, but the lower phase was the better preserved. The plan is simple, comprising a single room with benches along two wall faces and a narrow doorway in the top north-east corner. In the middle of one end of the room is a free-standing brick structure with reveals, directly in front of which a carefully delineated area of flooring is stepped down and partly burnt. The free-standing structure proved to be a hollow container and from within it came a large collection of clay sealings with cylinder seal impressions. Under this structure and set into its floor was an unbaked clay object of unusual shape and unknown purpose. Its fabric comprised large voids suggesting the clay had been carefully kneaded around some organic substance which has since decayed and disappeared.

This building appears to be a temple of Ninevite 5 date and the free-standing brick structure may be an altar or offering table with the recess in the floor being a receptacle for liquids, perhaps blood from offered animals. The date of the building is secure - from its floors and that of the overlying rebuild came quantities of incised and excised pottery dating to the later part of the Ninevite 5 period. Several AMS dates put the building at about 2,700 B.C. The closest comparison comes from the site of Kashkashok, less than 50 km to the west of Tell Brak, where a single-roomed structure with benches along two sides and with possible cult statues has been excavated by Syrian archaeologists. A similar plan, with bent-axis approach, also occurs at Tell Mozan in Temple BA, but that building is much larger and about 200 years later.

The collection of sealings from the HS4 building is of considerable interest. Only a handful of different seals are attested by the obverse seal impressions, most being of a curvilinear design. Crude animal scenes are also featured. If we turn the sealings over and examine their reverse faces we see that they had been originally affixed to a range of objects including possible box lids, often with traces of a curving peg set into a smooth riveted plaque or flange. This may represent an elaborate fastening for a box lid or for a removable cover for the altar itself. Along the plastered benches on the inner wall faces, there was evidence for the emplacement of pots and other containers. It may be that valued commodities, perhaps deposited within the room by temple devotees, were kept within these containers and sealed, either by the worshippers or by temple officials, to be re-opened under proper authority and then used as temple offerings on the altar.

Other Ninevite 5 Levels: Further Ninevite 5 levels have been briefly investigated by us at widely separated points of the mound. In Area HL burnt structures with typical pottery, including an elaborately decorated vessel, were explored. In Area HF at the north-east corner of the mound, a series of structures of late Ninevite 5 date included a room with brick fittings, partially bitumened floor and scatters of bird bones, which may have functioned as a kitchen. Another structure had a deeply stratified series of clean plaster floors. The main point of these investigations has been to demonstrate the extent of the settlement at Tell Brak during the Ninevite 5 period. There can be no doubt that during these centuries Tell Brak was one of the most important settlements in northern Mesopotamia, and this fact helps us to understand how Tell Brak may have quite smoothly been transformed from a major regional focus in the early-mid 3^rd millennium into an imperial administrative centre by the last quarter of that millennium.

Later 3^rd Millennium / Akkadian Period: Current work at the neighbouring sites of Tell Beydar, Tell Leilan and Tell Mozan is transforming our understanding of this region of northern Mesopotamia during the later 3^rd millennium. It is now clear that major developments in the complex processes of urbanisation and cultural fluorescence occurred before the imposition of south Mesopotamian Akkadian rule around 2,400 B.C. There is clearly a social and economic trend starting around the middle of the 3^rd millennium, or earlier, culminating in the establishment of Tell Brak as a major focus for Akkadian rule.

Micromorphology has shed valuable light on the deposition of this hoard by analysis of samples extracted from a section through the lane adjacent to the hoard building. Trampled deposits, rich in decayed plant remains, are overlain by a deposit of sterile wind-laid deposits and collapsed wall plaster and pisé fragments suggesting the decay into the lane of the adjacent building. Occupation then resumes after a spell of uncertain duration. This sequence of deposits puts the failure of the hoard depositor to retrieve their possessions into an informative context, indicating a spell of abandonment, at least in this part of the site, immediately subsequent to the deposition of the hoard.

Further illustration of the wealth and administrative power of Brak at this time is demonstrated by our excavations in Area HP at the south-west corner of the site. Here we exposed a large brick structure with associated rooms, yielding pottery and a rare inscribed piece of clay. Adjacent to the rooms was a large rubbish dump from which came hundreds of clay sealings with very fine cylinder seal impressions, showing a range of detailed and well-modelled scenes, closely paralleled by examples recently found at Tell Beydar. Functionally, these sealings had been used to seal door pegs on store-room doors. In sum, they indicate extensive administrative activity in this area, close to some of the massive public buildings previously excavated in Area SS by David Oates.

Post-Akkadian Period: Following the collapse of Akkadian rule in the north, the picture becomes less clear. Some scholars are suggesting a period of extensive regional turbulence and abandonment in the centuries around the end of the 3^rd millennium, but the evidence from Tell Brak remains unclear in detail. Partly in order to address this issue, we opened a trench in Area HN, not far from the summit of the mound, where we exposed levels dating to the first half of the 2^nd millennium. A building with a large central room, including plastered holes in the wall faces above floor level, was excavated and a sounding below this building uncovered traces of earlier levels, all with painted pottery. In the end, time ran out and we were unable chronologically to connect these lower levels, dating to the earlier 2^nd millennium, with our late HS levels dating to the very late 3^rd millennium. The nature of settlement at Tell Brak during these important centuries remains largely unknown.

Bibliography

Mallowan, M. E. L. 1947, 'Excavations at Brak and Chagar Bazar' Iraq, vol. 9, pp. 1-259.

Matthews, R. J. 1994, 'Imperial catastrophe or local incident? An Akkadian hoard from Tell Brak, Syria' Cambridge Archaeological Journal, vol. 4, pp. 290-302.

Oates, D., Oates. J. & McDonald, H. 1997, Excavations at Tell Brak. Vol. 1: The Mitanni and Old Babylonian Periods. Cambridge: McDonald Institute for Archaeological Research & British School of Archaeology in Iraq.

List of contexts