Introduction

Figure 1.

Location of Wadi Hammeh 27 in the east Jordan Valley.

The Natufian period (13 000–10 300 cal BC) in the Levant is considered a crucial juncture in human settlement history because it marks the transition between the mobile hunter-gatherers of earlier Epipalaeolithic phases (20 000–13 000 cal BC) and the sedentary, agrarian villagers of the Early Neolithic (10 300–8300 cal BC; Reference Bar-YosefBar-Yosef 2011). Wadi Hammeh 27 (c. 12 000–12 500 cal BC) in Jordan (Figure 1) is one of only a few known Early Natufian open-air, ‘base-camp’ type settlements (Bar-Yosef & Goren 1973: 67; Reference HenryHenry 1989: 219). It features large and complex pre-Neolithic buildings, a series of well-preserved artefact caches and activity areas, and a rich corpus of art. It is representative of the first sites where communities not only invested considerable energy in building residences in the form of substantial huts on stone foundations, but also refurbished and revisited them for generations—over hundreds and even thousands of years. The Natufian has been claimed as an example of pre-agricultural sedentism, but the length and frequency of its habitations remain unclear. One issue is that, for the majority of sites, long-term occupation of a single locale by hunter-gatherers would deplete food resources (cf. Reference MunroMunro 2004).

These concerns are the focus of a new La Trobe University project (Reference EdwardsEdwards 2014) entitled ‘Ice Age villagers of the Levant: sedentism and social connections in the Natufian period’, directed by the author and co-directed by Louise Shewan (Monash University/University of Warwick) and John Webb (La Trobe University). In order to achieve the project’s aims, the new excavations are intent on stripping away more of the overlying deposits of phases 2 and 3 at Wadi Hammeh 27 to expose the basal travertine layer (phase 4), where human burials are situated in rock-cut pits (Webb & Reference Edwards and EdwardsEdwards 2013). Through the use of strontium analysis of human teeth to estimate the length of residential occupations and the frequency of residential shifts, the project seeks to advance our understanding of how Natufian communities founded the earliest villages in the Jordan Valley around 12 500 cal BC.

The first series of excavations, conducted in the 1980s, focused on the site’s uppermost deposits in phase 1 (Reference Edwards and EdwardsEdwards 2013). A small sounding (XX F sondage) made at that time also demonstrated occupational continuity between the superimposed phases and the community memory of a sub-site burial by the building of successive cairns and other markers. These practices are among the first expressions of household tenure in the Middle East, or indeed anywhere in the world. Reference EarleEarle (2000) has suggested that the creation of boundaries in architectural space, including the internal demarcation of buildings by walls and other features, marks one of the indicators of property tenure in the archaeological record.

The 2014 excavations

Figure 2.

Aerial view of phase 2 occupation surface (plot XX F) excavated at Wadi Hammeh 27 in 2014; each grid square measures 1 × 1m; the earlier ‘XX F sondage’ (phase 4) is located to the right, and a pit dug from the later phase 1 is indicated by a dashed line; white discs indicate the locations of ‘artefact clusters’.

The clearance in 2014 of the phase 2 occupation at Wadi Hammeh 27 under structure 1 (Figure 2) brought to light further examples of inter-generational continuity. For example, feature 7 (square C3), a double-coursed stone ring, is a predecessor to a version of the same feature that continued in use in the overlying phase 1 (Reference Edwards and EdwardsEdwards 2013: 73). Feature 4 (square D5) is an oblong stone platform that extends east to square D4 and south into squares C4 and C5. It is truncated by the pit in the north-east corner of square D5. A flat stone is placed at the centre of the stone group, and the feature may have functioned as a post-support, as did several similar features in phase 1 (Reference Edwards and EdwardsEdwards 2013: 71). The pit that cuts the feature was itself capped by a roughly rectangular stone platform in phase 1. Square E6 yielded a dense agglomeration of finds (Figure 3), including a zoomorphic basaltic pestle (RN 140049), a phalliform figurine (RN 140225), a basaltic shaft-straightener (RN 140047) and a basaltic handstone (RN 140048). The clustering of materials here reflects a similar concentration in the overlying phase 1 near the light, where structure 1 opened to the west.

The first series of excavations yielded numerous artefact clusters and caches of raw materials (Edwards & Hardy-Smith 2013: 105). Artefact clusters 1–17 occurred in phase 1 contexts, and artefact cluster 18 was found in phase 2 (in the plot XX F sondage). The recent season yielded three additional clusters on the phase 2 floor (Figure 2). Artefact cluster 19 comprised a pair of large flint blades nearly 10cm long, an unusually large size for Wadi Hammeh 27; artefact cluster 20 consisted of three lightly reduced and apparently heat-treated flint cores, and two limestone cobbles; artefact cluster 21 (Figure 3) included a cache of 138 dentalium (Antalis sp.) fragments overlying a pair of retouched blade tools, including a Helwan retouched awl.

Figure 3.

View over square E6.1, indicating artefact cluster 21—a cache of dentalium fragments—in the lower right; farther towards the right-hand corner is a zoomorphic basaltic pestle (RN 140049); then proceeding clockwise, a basaltic handstone (RN 140048), a phalliform figurine (RN 140225) and a basaltic shaft-straightener (RN 140047).

Figure 4.

1) Zoomorphic basaltic pestle (RN 140049); 2) phalliform basaltic pestle (RN140053).

New art pieces

The 2014 season also produced several important artefacts, including some with close parallels to items found at the dawn of research into the Natufian culture nearly a century ago. Of the basaltic artefacts, the most notable finds were two diminutive items (Figure 4); the first, a zoomorphic pestle (RN140049), and the second, a phalliform basaltic pestle (RN140053). The key decorative features of RN 140049 are a raised band circumscribing the shaft near the distal end and an obliquely shaped terminus. In combination, the features suggest an ungulate hoof. The piece finds several Early Natufian parallels in objects from El Wad Cave at Mount Carmel (Garrod & Bate 1937: pl. XV.4), and Hayonim Cave in western Galilee (Reference Belfer-Cohen, Bar-Yosef and VallaBelfer-Cohen 1991: fig. 7: 1, 3 & 5). The distal end of RN 140053 features a raised band circumscribing the shaft and at its end the piece is lightly grooved, conveying the impression of a phallic object. The rich square, E6, also yielded a large phalliform limestone figurine (RN 140225). The piece was complete but cracked in situ, and repaired as three conjoinable fragments (Figure 5). A white colouration caps the proximal end of the piece, whereas the rest of the object is light to dark grey. In order to highlight the proximal cap, an area of the piece down the right lateral margin had been reduced by pecking, emphasising the white terminus above it.

Figure 5.

Phalliform limestone figurine (RN 140225).

Figure 6.

Carved and incised bone animal head (RN 140226).

The most distinctive find of the season, also from square E6, was a carved and incised bone animal head (RN 140226; Figure 6). It is evidently the terminal piece of a larger object such as a sickle. Although badly damaged and calcined by burning, the rendering of an ungulate animal, probably a gazelle, is compelling. The piece has parallels with objects from Kebara Cave (Reference Turville-PetreTurville-Petre 1932) and El Wad Cave (Garrod & Bate 1937: pl. XIII.3). Together, the animal head and the zoomorphic pestle strengthen the links between Wadi Hammeh 27, lying east of the Jordan, and key Natufian settlements to its west.