Introduction

Figure 1.

Wadi Kubbaniya, the Late Paleolithic dune field and WK26.

In 2014, members of the Combined Prehistoric Expedition Foundation and the Aswan-Kom Ombo Archaeological Project (CPEF/AKAP) undertook investigations at site WK26 in Wadi Kubbaniya, Egypt. The site consists of a lithic scatter on the west side of the wadi, across from the Late Palaeolithic dune field explored by the Combined Prehistoric Expedition between 1978 and 1983 (Wendorf et al. 1980, 1986, 1989) (Figure 1). Based on radiocarbon dates and its stratigraphic position, WK26 dates to the end of the Late Palaeolithic Kubbaniya sequence (Wendorf & Schild 1989). Although sites of the early and middle portions of the sequence are well-documented, few sites dating to the end of the sequence have been identified and explored in the wadi or elsewhere in this area of Upper Egypt. The presence of hearths, postholes and depressions, along with faunal and floral remains, provide an insight into settlement and subsistence at the end of the Late Palaeolithic in the wadi.

Figure 2.

WK26.

WK26 was first recorded in 2012 (Figure 2). An ashy area on top of apparent playa silts yielded a radiocarbon date of 12 060 BP +/- 50 RCYBP (Beta-319442). In addition, an Ounan point was found on the surface; such points have been dated elsewhere to the early-mid Holocene (Wendorf et al. 2001; Reference Macdonald and HopeMacDonald 2003; Riemer et al. 2004; Cancellieri & di Lernia 2014). These factors suggested that WK26 dated to the end of the Late Palaeolithic Kubbaniya sequence and was possibly more recent. Due to the potential age of the site, the CPEF/AKAP undertook excavations.

Excavations at WK26

The site is at a higher elevation than the Late Palaeolithic dune field and is associated with playa silts, which were interpreted as the Upper Kubbaniya Silts (Wendorf & Schild 1989). Two areas—A and B/C—were excavated, both to a depth of approximately 30cm. The profile in an exploratory trench that bordered area A indicated the possible presence of two underlying, thin cultural horizons. Area A, which measured 10 × 10m, contained 27 features: 2 hearths, 3 ashy areas, 1 stained area, 1 natural depression, the remnants of a pit, and 19 post holes (Figure 3). Area B/C, which measured 10 × 17m, produced 15 features, most of which were ashy areas and the remnants of apparent storage features (Figure 4). No other sites within the wadi have produced a similar number and diversity of features.

Figure 3.

WK26 Area A.

Figure 4.

WK26 Area B.

The hearths and post moulds in area A are embedded in playa silts and these silts distinguish the stratigraphy here. The features in area B appear to be below playa silts; only traces of silts were found here. Charcoal from the hearths and two features in area B/C was dated to 13100 ± 35 RCYBP (PRI-14-041-1), 13478 ± 35 RCYBP (PRI-14-041-2) and 13553 ± 34 RCYBP (PRI-14-041-3), placing this site at the end of the Late Palaeolithic sequence and possibly associating it with high stands of the ‘wild Nile’ (Reference ButzerButzer 1997).

Figure 5.

WK26 artefacts: isosceles triangles, top row; backed elements with Ouchtata retouch, middle row; endsrapers, bottom row.

The lithic technology is simple. The assemblage is flake-oriented and revolves around the removal of blanks from single platform cores. Chert and Egyptian flint are the predominant raw materials. The retouched tools differ from those from the sites elsewhere within the wadi. The assemblages consist of backed and Ouchtata elements, truncations, endscrapers and geometric microliths. The most distinctive tool type that sets the assemblage apart from sites across the wadi is a backed and truncated piece that resembles a large isosceles triangle (Figure 5). These pieces are almost exclusively backed on the right side and truncated at the proximal end.

The assemblage also includes grinding implements—handstones and grinding stones—but these, especially the stones, differ from those found in the dune field. The grinding stones generally tend to be thin slabs with shallow but distinct grinding surfaces rather than the large globular blocks in the dune field. The handstones are circular, often with two grinding surfaces. Several of the handstones and slabs have traces of ochre.

Faunal remains consist almost exclusively of fish; no remains of large mammals were found. Most of the fish remains consist of Clarias. This predominance of fish remains further distinguishes WK26 from the other Late Palaeolithic sites within the wadi. The hearth remains and post moulds within the silts and the dominance of fish indicate that the site represents a dry season occupation, probably exploiting fish caught in pools. The apparent paucity of contemporary sites implies that the wadi was only occupied seasonally, possibly only occasionally, and was part of a settlement round that extended outside the wadi.

As a proof of concept, grinding faces on six implements were washed to recover pollen, phytoliths and starch. Four were found ‘face down’ on the site surface and two were from the subsurface. Pollen from members of the mustard and amaranth families suggests seed processing at the site. Phytoliths typical of festucoid grasses predominate, followed by a few chloridoid and panicoid phytoliths. Two of the four stones from the surface yielded phytoliths that indicate cutting by a threshing sledge or trampling; these surface stones could not be dated. Several ‘modified’ sheet phytoliths had been burned, indicating the parching of grass seed or cereal prior to grinding. One grinding stone also contained a phytolith with torn edges and was typical of stems that have been cut with sickles. Another grinding stone that did not contain dendritic phytoliths displayed an echinate spherical phytolith from a palm suggesting the grinding of dates.

Figure 6.

Plant element from grinding implement.

One of the handstones recovered from the Late Palaeolithic level yielded phytolith sheet elements exhibiting cuts suggestive of post-harvest processing (Figure 6). These Poaceae sheet elements are still under study. The second grinding stone from this level had ochre but few phytoliths, which indicates that it was used for grinding something other than floral items.

WK26 further underscores the settlement and subsistence complexity and diversity of the Late Palaeolithic in this portion of Nubia/Upper Egypt. The wadi is the only place in this region investigated to date where this diversity of activity is exemplified to such an extent. Although most of the sites here are associated with the Kubbaniyan industry, other sites are related to a greater or lesser extent with another five industries. The differences in the lithic typology and technology, including grinding implements, demonstrate that the Late Palaeolithic was a period of diversification and regional differentiation.