In 1983, the investigations of the Combined Prehistoric Expedition at 27 locales at Wadi Kubbaniya in Upper Egypt were concluded after four field seasons. This work culminated in four comprehensive publications highlighting the importance of the Kubbaniyan lithic industry during the Late Palaeolithic (Wendorf et al. 1980, 1986, 1989a & b). Wadi Kubbaniya is located north of Aswan and is the largest wadi in the Western Desert of Upper Egypt. During the Late Palaeolithic, overflow from the Nile became impounded in the wadi, forming a lake. An extensive dunefield formed along the north-eastern edge of this lake; Late Palaeolithic people repeatedly camped within and adjacent to this dunefield (Figure 1). This presence dates from about 20 000 BP to around 12 000 BP. The length and intensity of occupation varied but, based on the variety of artefacts, abundant faunal remains, hearths and ash lenses, and numerous grinding implements, most loci of activity appear to have been domestic occupations. The grinding implements are evidence that plant resources were an integral aspect of subsistence.
In 2012, the Combined Prehistoric Expedition Foundation and Yale University-University of Bologna Aswan-Kom Ombo Archaeological Project renewed investigations in the wadi, discovering and investigating site WK25. This is a small lithic concentration situated on a hummock in the middle of the dunefield (Figure 2) between a number of previously investigated sites (E-78-2, E-78-3 and E-78-4), some of which contained deep deposits and many features and artefacts, representing the most intensively occupied loci known in the wadi (E-78-3, E-78-4 and E-81-1).
At WK25, 25m² were excavated to a depth of 0.6m (Figure 3). The profile consists of unconsolidated sand with little internal stratigraphy (Figure 4). Faunal remains were sparse, represented by slivers and fragments. A preliminary examination indicates that most remains are fish; identifiable fragments are Nile catfish (Clarias).
The assemblage contains 3452 debitage pieces, 168 cores and 98 retouched tools. Flakes predominate among the debitage. Like other Kubbaniyan assemblages, backed elements, especially partially backed bladelets, dominate the retouched tools (Table 1). Nucleiform endscrapers, however, which are rare to non-existent in other assemblages in the wadi, are a distinctive component (Figure 5). These scrapers resemble cores but with additional retouch along one or more edges. Cores are dominated by the single platform variety. Grinding implements typical of other loci are lacking.
The predominant raw material is chert, with Egyptian flint a distant second. The source of this chert is unknown. Gravel sheets near the dunefield and lenses in the profile at a bir in the wadi contain cobbles of similar dimensions as the initially struck cores. Most cobbles in the wadi, however, are quartz, quartzitic sandstone or unknappable materials; suitable chert nodules are rare. Cherts do not appear to have been procured locally within the wadi, though initially struck cores suggest that chert was available sufficiently close that cobbles were brought to WK25.
|Table 1. Retouched tools|
|4. Rabot (nucleiform endscraper)||9|
|5. Denticulated endscraper||2|
|12. Simple perforator||1|
|17. Dihedral burin||1|
|18. Dihedral angle burin||1|
|19. Angle burin on a break||2|
|22. Angle burin on a straight, oblique truncation||1|
|28. Nuclieform burin||1|
|45. Straight-backed pointed bladelet||1|
|55. Arch-ended bladelet||3|
|56. Arch-backed bladelet||6|
|63. Partially backed bladelet||25|
|66. Backed bladelet fragments||3|
|74. Notched flake||1|
|75. Denticulated flake||1|
|76. Notched blade||3|
|79. Notched/denticulated piece w/ retouch||2|
|80. Truncated piece||2|
|105. Piece with continuous retouch||29|
The outstanding feature of the assemblage is that cores outnumber tools 1.7:1. Lengths of single platform cores have a bimodal distribution, with peaks at 20–21mm and 25mm. 62 cores are under 23mm long and over half of these are less than 16mm thick. The larger cores tend to be globular, while the smaller cores tend to be rectilinear in plan and lenticular in cross-section (Figure 6). Most cores, especially the smaller ones, appear to be 'played out', suggesting that these cores were curated and used to the maximum extent, underscoring that chert was not obtained locally.
An ASM conventional radiocarbon date of 17 870±70 BP (Beta 319441) places WK25 early in the Late Palaeolithic sequence, contemporaneous with E-78-3, E-78.4 and E-81-1. Six factors, however, distinguish WK25:
These factors suggest that WK25 was an activity-specific lithic workshop, rather than a domestic locus. As such, WK25 represents a previously unidentified aspect of the Late Palaeolithic Kubbaniyan. If the entire dunefield is considered to be a 'site', then it appears to contain loci of intense and repeated occupations where a variety of domestic activities occurred, such as E-78-3, E-78-4 and E-81-1, as well as discrete and localised activity-specific loci, such as WK25. This pattern appears to have extended to areas adjacent to the dunefield.
Wadi Kubbaniya represents one of the most, if not the most, intensively investigated and reported Late Palaeolithic occupations in Upper Egypt and Nubia. The wadi illustrates the diversity and complexity of the Late Palaeolithic period in the Nile Valley both with respect to the variety of industries and settlement-subsistence strategies. The composition of E-78-3, E-78-4 and E-81-1 suggests occupation here was most likely part of a seasonal round that extended beyond the wadi. WK25 further contributes to our understanding of the nature and complexity of such Late Palaeolithic activity in Upper Egypt and Nubia.