Jörg Linstädter (Cologne, Germany): Systems of prehistoric land use in the Gilf Kebir

In: HAWASS, Z. (ed.) Egyptology at the dawn of the Twenty-first century. Proceedings of the Eigth International Congress of Egyptologists, Cairo 2000, American University of Cairo Press. Cairo 2002.

Introduction

The Gilf Kebir is a sandstone plateau situated in the margin of south-western Egypt. On its south-eastern fringes broad Wadis flow into the plains. Over the period of the past 20 years their upper reaches have been subject to archaeological research of scientists of the University of Cologne (KUPER 1995, SCHÖN 1994).

Considering the history of ancient Egypt and especially its pharaonic period, the Gilf Kebir seems to be completely marginal. There are up to now no confirmed data available on the existence of direct or indirect relations between the Gilf Kebir area and the Nile valley in prehistoric times. But not only historical events can connect egyptology and prehistoric archaeology, but also shared methodological approaches. Modern archaeology is more than a descriptive historical science. Scientific branches have been developed such as economical, sociological and ecological archaeology in order to reconstruct systems of land use, of mobility, and cognition. The term landscape archaeology has been coined to summarise the scope of these different branches, which are relevant for all archaeological and cultural anthropological disciplines (LÜNING 1997:277). This paper will try to show the potential impact of this approach on the understanding of archaeological sites in the study area.

Study area

The plateau of the Gilf Kebir is situated 650 km west of the Nile valley on the same geographical latitude as the Aswan lake. In the North it disappears under the Great Sand Sea, while in the South its steep cliffs rise about 150 m above the surrounding plains. Here the plateau is intersected by broad wadis, one of them is the Wadi Bakht (fig. 1). It is of special interest for geographers and archaeologists because of the unique geomorphological situation of the so-called barrier dune (KRÖPELIN 1987:286). Nineteen kilometres above the wadi mouth, this barrier dune blocks the wadi which is flowing in an easterly direction.

fig.1: Sites of the lower reaches and the eastern plains.

The dune sands come exclusively from the Great Sand Sea north of the Gilf Kebir. The remains of the barrier dune visible today consist of a late pleistocene dune, topped by a recent leedune. Over a period of several millenniums, the sediments of a playa were accumulated behind this barrier. The lacustrine sediments cover an area of 6600 m². On one occasion, about 3,800 calBC, the playa sediments were too thick and the dam collapsed after an event of heavy rainfall (KRÖPELIN 1989). The break is located in the middle of the barrier and allows a view into a sediment body more than eight metres high.

In connection with archaeozoological and archaeobotanical investigations (GAUTIER 1982, NEUMANN 1989, PETERS 1987), the analyses of the section consisting of clay and sand layers rendered some hints for the reconstruction of the climatic changes in the holocene. In spite of certain differences of the results, all scientists agree with the assumption that even in times with favourable climate, the conditions for human life were little better than today. Annual precipitation should not have been higher than 150 mm (KRÖPELIN 1989). Therefore, only the special topographical situation of the Wadi Bakht is responsible for the comparably high species diversity of flora and fauna and the possibility of human land use which existed over a period of several centuries.

What has already been done?

All archaeological sites documented up to now can be fit into a period of time between 8,000 and 3,700 BC characterised by comparably favourable climatic conditions (HALLIER 1996, LINSTÄDTER 1999). Investigations in this part of the valley already started in the thirties of this century and continued till today.

All field investigations, however, were limited to the area close to the barrier dune because of its good accessibility and the wealth of archaeological material found there (BAGNOLD et al. 1939, McHUGH 1982, WENDORF & SCHILD 1980). The surrounding areas of the plateau and the Eastern plains have never been subject to a closer observation. The accumulation of finds in the valley led to the impression that the mountains were not suitable for human settlement. Therefore the valleys with their water resources were interpreted as a place of refuge for prehistoric people.

We used new approaches of our current research project titled “ACACIA – Arid Climate, Adaptation and Cultural Innovation in Africa” as an opportunity for further examinations of the finds excavated by the BOS project in the early eighties.

The occupation of this part of the valley can be roughly divided into two phases named middle and late neolithic in congruence with the common nomenclature of the Eastern Sahara (WENDORF et al. 1984). The material culture is dominated by stone artefacts. The middle neolithic, lasting from about 5,500 to 4,350 BC, is characterised by a highly specialised blade production and a differentiated tool kit. The stone artefacts are accompanied by thick undecorated pottery, ostrich egg shell artefacts and ground stone implements (LINSTÄDTER 1999).

The subsequent phase of the late neolithic, from 4,300 to 3,500 BC, is dominated by a roughly worked flake industry with no diagnostic traits, opposed by a fine decorated pottery (HALLIER 1996). Additional finds, especially bones, probably indicate a hunter-gatherer population in the middle neolithic and a mainly pastoral nomadic population in the late neolithic.

Expectations

A vast middle neolithic inventory from the barrier dune’s surface allowed, besides typological investigations, the reconstruction of a production sequence, describing the process of stone tool production in the course from the acquisition of raw material to the tool completion (fig.2) (LINSTÄDTER 1999). The integration of the lithic industry into the reconstruction of the holistic picture of prehistoric life leads to further questions. They are mostly linked to foraging, communication and funeral rites.

fig.2: Production sequence of a Middle Neolithic inventory.

Geological investigations carried out on the plateau led to the assumption that exploitable deposits of quarzite could be found. The described part of the valley is definitely too small to practise all economic activities indicated by archaeological records, and so the lower reaches and the eastern plains of the Wadi Bakht were considered for these activities.

The use of land use concepts

There are certain theoretical implications of the term landscape archaeology and how it could help with the interpretation of field data (McGLADE 1999).

By turning away from the big settlements, we also turn away from a more traditional approach: The excavation of central places and the investigation of their history. We go out into the landscape to look for economic areas, for raw material sources, and trails (DARVILL 1999). What we hope to find are small, specialised, quite often unspectacular sites which could help us to answer questions about settlement structure, mobility and economic exploitation of landscapes.

Concepts of land use can be similar in different times and at different places, because they are influenced by the adaptation to environmental factors like climate and topography. On the other hand, they bear individual traits of human populations. Human populations are not completely forced to react on environmental changes. They have also a specific pool of concepts, allowing them (within a distinctive frame) to choose an individual strategy of subsistence.

In these cases concepts of land use can contribute to the subdivision of the history of occupation, if conventional methods like stratigraphy, typology and absolute dating are not sufficient. When we find a workshop with preforms of a certain raw material and we find a camp site with blanks on the base of these preform and from the same raw material, we can connect them and assume simultaneity - an information which Radiocarbon dating could never give.

Which advantages has the Gilf Kebir?

The Gilf Kebir is highly suitable for our approach because of two reasons. The first is climate. Due to the mere absence of precipitation, water as the most important factor for shaping the topography has no significance here. In our study area, nothing has changed the surface since the so-called Neolithic Wet Phase. Even in the lower reaches of the Wadi Bakht and on the eastern plains of the Gilf Kebir which are both mainly characterised by fluviatile sediments, many sites were found on alluvial soils and shallow terraces.

Another advantage of the arid climate is the good accessibility of sites. But this also implies some disadvantages for the archaeologist. As soils were not formed or were eroded due to the lack of a vegetation cover, all finds are now accumulated on the surface. Furthermore, artefacts were exposed to the destructive power of the harsh climate over several millenniums. Organic pieces and ceramic could offer the least resistance. This situation facilitates a large-scale mapping, but implies problems with the dating of the material.

The second factor why the Gilf Kebir is highly suitable for our approach is man himself. Due to today’s precipitation regime no independent human life can be sustained in the Gilf Kebir. Building and agriculture which cover and destroy archaeological sites in Europe or in the Nile valley are completely absent. This has again beneficial implications on the accessibility of sites.

What is the relevance of landscape archaeology?

In the following some aspects of landscape archaeology, relevant for the investigations in the Gilf Kebir, will be summarised. Although closely related to the classical archaeology of settlements, the landscape archaeology emphasises the “in between” of the big settlements. A habitat is not restricted to the settlement, but its surrounding, functioning as an economic or a traffic area. In traditional approaches, landscape was regarded as a scenery or background for human activities which is regulated by its own laws. An archaeology of landscape tries to unite knowledge from natural and cultural sciences with the aim to create a dynamic system (TILLEY 1994). In this system landscape becomes a sphere not completely explainable by natural sciences but created by human cognition and agency and therefore being an holistic entity.

Egyptology also deals with this subject. The geomorphology of the Nile valley and its economic exploitation in pharaonic times are very suitable for an approach of landscape archaeology. In our interdisciplinary project at the University of Cologne, this connection between geomorphology and exploitation is subject to egyptological investigations (HERB 2000). The results indicate that certain land forms such as floodplains, geziras, permanent swamps and the adjacent desert are closely connected to certain economic activities such as agriculture, horticulture, keeping lifestock, and hunting. An extraordinary result is that the perception of land use systems even influenced the concepts of wall decoration in the tombs.

Survey of the plateau and the Eastern plains in 1999

To gain further information on raw material acquirement, economic areas, mobility etc. we undertook a survey of the plateau and the Eastern plains in 1999.

The area of the plains stretching over several km² was surveyed by car. We mainly hoped to detect traces of a neolithic use of this region for hunting and grazing. The surface in the survey area is dominated by fluviatile sediments. This implicates a high surface instability and normally a bad preservation of sites. In spite of this, we found twenty-six sites of very different size (fig.1).

Later the plateau close to the barrier dune was surveyed by foot on an area of approximately 20 km². In this case the main aim was to search for raw material sources and adjacent workshops. As mentioned above, the first steps of the proposed production sequence are thought to have been performed here.

Results of the survey

In the plains and the lower reaches of the wadi were found both smaller sites (sometimes with stone circles), and camp sites, marked on the map (fig.3) by an open circle. These large areas with find accumulations always included groundstone implements. Only in the plains isolated workshops were found, and ateliers close to quarzite outcrops.

fig.3: Sites of the plateau survey 1999.

During the Plateau survey we found ample raw material sources and traces of preform processing (fig.3). Furthermore the occurrence of merely all varieties of quarzite and basalt used down in the valley was proofed.

In addition to our expectations, the ateliers bore artefacts which indicate more activity than preform production. Quite often ateliers were accompanied by proper camp sites. The appearance of pottery, grindstone implements and house constructions suggests a longer stay. In the map (fig.3) the sites are represented by symbols of different tool types typical for the sites. Workshops and camp sites are easy to differentiate. The first ones are characterised by cores and flakes. The northern and southern rim of the plateau for example is lined by such workshops. On the other hand, at the camp sites (marked by grey patches) pottery, grindstone implements and other indicators for settlements were found.

Three camp sites deserve special attention. On the map they are named boulder hill, basalt hill and stone circles. They are situated on convenient positions close to the valley and have several conditions in common: the availability of water and raw material as well as a good view. Remarkably, all sites are characterised by stone constructions. A range of buildings can be distinguished, reaching from simple wind shelters and circular constructions to complex dwellings consisting of several units. Up to now, we only had little time for the documentation of each site. More detailed investigations and excavations are needed in order to interprete their function.

Discussion

We have discovered an unknown prehistoric settlement area, integrating the already documented sites from the valley. Its extension and its history of occupation and exploitation is still to be investigated. Our finds excelled our expectations by far. Instead of small stockposts expected in the plains and workshops expected on the plateau we found in both cases large settlements, being themselves centres of further economic activity. Certainly our image of the Gilf Kebir, where the Wadis are refuges only for bad times, has to be reconsidered. For sure the barrier dune was a place highly frequented by prehistoric people. The thesis of a concentric settlement system with the dune as a centre and the plateau as a mere periphery is, at least for the late neolithic, not longer tenable.

fig.4: Proposed land use systems.

A chronological positioning of the sites is problematic. Samples for absolute dating can only be obtained in excavations. Artefacts suitable for a determination by typological comparisons are rare. For example, the typical middle neolithic inventory of a differentiated tool kit was not found.

On the other hand, useful results can be derived from the concept of a production sequence. All artefact types found at the workshops can be assigned to the proposed acquisition phase. They are obviously preforms to the blank production reconstructed for the valley. The completion of the reduction sequence is one of the most remarkable results of our survey.

Pottery was rarely found and is exclusively from late neolithic times. Potsherds only appear on camp sites together with other indicators for a settlement.

This might be a first step in the reconstruction of different land use systems (fig.4). Was the only activity of middle neolithic hunter-gatherers on the plateau the exploitation of raw material sources close to the plateau rim? In contrast to this, did their descendants from late neolithic times, who probably kept livestock, integrate the plateau in all economic activities?

In summary it is safe to say that at least two complex land use systems appear, consisting of a network of hierarchical sites. The specific function of the sites is yet to investigate.

References

BAGNOLD, R.A., O.H. MYERS, R.F. PEEL & H.A. WINKLER (1939) An expedition to the Gilf Kebir and `Uweinat 1938. The Geographical Journal 93, 1939: 281-313.

DARVILL, T. (1999) The historic environment, historic landscapes and space-time-action models in landscape archaeology. UCKO & LAYTON 1999: 104-118.

EL-BAZ, F & M.A. MAXWELL (eds.) (1982) Desert landforms in Southwest-Egypt: A basis for comparison with Mars. Washinton 1982.

GAUTIER, A. (1982) Neolithic faunal remains in the Gilf Kebir and the Abu Hussein Dunefield, Western Desert, Egypt. In: EL-BAZ, F & M.A. MAXWELL (eds.) (1982): 335-339.

HALLIER, M. (1996) Zwei keramische Fundplätze am Übergang vom 5. zum 4. Jahrtausend vor Christi Geburt in Südwest-Ägypten: Wadi Bakht 82/15 und 82/24. (Magisterarbeit Köln 1996 unpublished).

HERB, M. (2000) Das Altägyptische “Fischerstechen”. Untersuchungen zu einem festlichen Wettkampf des dritten Jahrtausends vor Christus. Ph.D. unpub., Köln 2000: 333-557.

KRÖPELIN, S. (1987) Palaeoclimatic evidence from Early to Midholocene playas in the Gilf Kebir (Southwest-Egypt). Palaeoecology of Africa 18, 1987: 189-208.

KRÖPELIN, S. (1989) Untersuchungen zu Sedimentationsmilieu von Playas im Gilf Kebir (Südwest-Ägypten). In KUPER, Hrsg., 1989: 183-305.

KUPER, R. (ed.) (1989) Forschungen zur Umweltgeschichte der Ostsahara. Africa Praehistorica 2, Köln 1989.

KUPER, R. (1995) Prehistoric Research in the Southern Libyan Desert. A brief account and some conclusions of the B.O.S. projekt. Cahier de Recherches de L`Institut de Papyrologie et d`Egyptologie de Lille 19, 1995: 123-140.

LINSTÄDTER, J. (1999) Leben auf der Düne. Der mittelneolithische Siedlungsplatz Wadi Bakht 82/21 im Gilf Kebir (Südwest-Ägypten). Archäologische Informationen 22/1, 1999:115-124.

LÜNING, J. (1997) Landschaftsarchäologie in Deutschland – Ein Programm. Archäologisches Nachrichtenblatt 3: 277-285.

McGLADE, J. (1999) Archaeology and the evolution of cultural landscapes: towards an interdisciplinary research agenda. UCKO & LAYTON 1999: 458-482.

McHUGH, W.P. (1982) Archaeological Investigations in the Gilf Kebir and Abu Hussein Dunefield, Western Desert, Egypt. In EL-BAZ, F & M.A. MAXWELL (eds.) 1982: 301-334.

NEUMANN, K. (1989) Zur Vegetationsgeschichte der Ostsahara im Holozän. Holzkohlen aus prähistorischen Fundstellen. In KUPER, R. (ed.) 1989: 13-181.

PETERS, J. (1987) The faunal remains collected by the Bagnold-Mond Expedition in the Gilf Kebir and Gebel Uweinat in 1938. Archéologie du Nil moyen 2, 1987: 251-264.

SCHÖN, W (1994) The Late Neolithic of Wadi el Akhdar (Gilf Kebir) and the Eastern Sahara. Archéologie du Nil moyen 6, 1994: 131-175.

TILLEY, C. (1994) A phenomenology of landscape: places, paths and monuments. Berg, Oxford 1994.

UCKO, J. & R. LAYTON (eds.) (1999) The Archaeology and Anthropology of Landscape. Shaping your Landscape. Routledge; London & New York 1999.

WENDORF, F.& R. SCHILD (1980) Prehistory of the Eastern Sahara. New York 1980.

WENDORF, F., R. SCHILD & A. CLOSE (1984) Cattle keepers of the Eastern Sahara: The neolithic of Bir Kiseiba. Dallas 1984.