Dates for the Early Natufian of El-wad Terrace, Mount Carmel, Israel

The Natufian culture of the southern Levant played an integral role in the transition from simple huntergatherers to food-producing societies of the Neolithic, but the major Natufian hamlets are currently poorly dated. Moreover, none of these complex, continuously occupied base camps have delivered an adequate number of dates to enable an in-depth delineation of intra-Natufian developments. This paper presents the first results of our dating program at el-Wad terrace, Mount Carmel (Israel), one of the major Natufian hamlets of the “core area” of this culture. Thirteen accelerator mass spectrometry (AMS) radiocarbon age determinations were obtained from 46 bone (both ungulate and human) and charcoal samples, originating in Early Natufian living surfaces, dwellings, and burials. The obtained dates are largely in agreement with the cultural affiliation of the samples (13–15 kyr cal BP). Two series of dates from different locations show good agreement with the stratigraphy. The ages of the burials clearly point to their being younger than the living surfaces seemingly associated with them. Presently, no burials may be linked with the major architectural phase of Early Natufian el-Wad. Our ongoing dating program and the processing of additional samples from refined contexts will help shed important light on the initial phases of the Natufian culture, habitation duration, intensity, and continuity, as well as the relationships between site features and stratigraphy. INTRODUCTION The Natufian culture of the southern Levant played an integral role in the transition from simple hunter-gatherers to food-producing societies of the Neolithic (e.g. Garrod 1957; Henry 1995; Belfer-Cohen and Bar-Yosef 2000; Bar-Yosef 2002). In spite of its pivotal place in the prehistory of the Levant, sites of the Natufian culture are currently poorly dated (Weinstein-Evron 1998; Stutz 2004; Blockley and Pinhasi 2011; Maher et al. 2011). Moreover, the relatively scarce dates are derived from various sites and none of the complex, continuous base-camp sequences have delivered an adequate number of dates to enable an in-depth delineation of intra-Natufian developments. In this paper, we present the first results of our dating program at el-Wad terrace, Mount Carmel (Israel), one of the major Natufian hamlets of the “core area” of this culture. Based on its long Natufian sequence, encompassing all main stages of the culture, the rich material culture, architectural remains, and numerous and varied burials, el-Wad has long been regarded as one of the key Natufian sites since its first excavation in the early 20th century (Garrod and Bate 1937; Weinstein-Evron 2009). Today, following extensive archaeological and archival research, the site appears much more complex than previously thought, both chronologically and spatially. However, we still lack detailed information about the sequence of occupations, their nature, organization of space, permanency of settlement, and subsistence strategies and how they all may have changed through time, and need to establish the chronological framework for these changes. We expect our dating program to generate a more comprehensive portrayal of the settlement at this particular site and of the Natufian in general and to yield a new understanding of increasing sedentism and societal complexity at the threshold of the “Agricultural Revolution.” 1Zinman Institute of Archaeology, University of Haifa, Haifa 31905, Israel. 2Corresponding author. Email: [email protected]. 3Radiocarbon Dating and Cosmogenic Isotopes Laboratory, Kimmel Center of Archaeological Science, Weizmann Institute, 76100 Rehovot, Israel. 4University of Bonn, INRES-Soil Science, Nussallee 13, 53115 Bonn, Germany. 814 M Weinstein-Evron et al. EL-WAD: THE SITE AND EXCAVATION El-Wad is a large Natufian site in Mount Carmel, well known for its elaborate architectural features and numerous graves and high density of finds. The site (a large cave and adjacent terrace; Figure 1) contains a long and rich Early to Late/Final Natufian sequence. It was first investigated by C Lambert in 1928, but became well-known as a result of the 1929–1933 excavation campaign of D A E Garrod (Garrod and Bate 1937; see also Weinstein-Evron 2009). The site was later excavated by F Valla and O Bar-Yosef (Valla et al. 1986), by M Weinstein-Evron (1998), and recently by M Weinstein-Evron, D Kaufman, and R Yeshurun (Weinstein-Evron et al. 2007; in press; Yeshurun et al., in press). As documented previously (e.g. Garrod and Bate 1937; Garrod 1957), the Early Natufian layer is much thicker and richer than the Late Natufian. In recent years, during the renewed excavations on the northeastern part of the el-Wad terrace, the later phases of the Early Natufian have been exposed in an area of ~70 m2. Throughout the site, this is the phase displaying substantial stone architecture, as previously documented by Lambert and Garrod (Weinstein-Evron 2009; WeinsteinEvron et al., in press; Figure 1b). A well-built curvilinear wall (Wall I), minimally 9 m in diameter, is the most prominent feature of the current excavation on the terrace (Figure 1). This wall partially abuts a natural bedrock step in a manner similar to the “terrace wall” unearthed by Garrod (Garrod and Bate 1937: Plate V1-2). Wall I could have been used for leveling the naturally steep bedrock terrace and/or for demarcating a large structure. In the area enclosed by Wall I, several partially preserved stone structures and stone-rich “living floors” (which are not always clearly associated with walls) have been excavated. Among the former, Structure II (Figures 1, 2) is the best preserved example of a medium-sized dwelling in the area enclosed by Wall I. Three successive stone-laid pavements were unearthed in it to date, thereby attesting to the repeated use of the structure. The density of finds in the Early Natufian layer is extremely high, but human remains are almost absent; so far, only 1 burial has been exposed. The stone structures, numerous living floors, density and diversity of finds, and the near-complete absence of burials indicate that this part of the site was used primarily for habitation and daily activities in the Late Early Natufian (LEN) (Weinstein-Evron et al., in press; Yeshurun et al., in press). To date, 3 radiocarbon dates have been obtained for the bottom, middle, and upper parts of the relatively shallow Natufian layer in the inner part of the cave (Weinstein-Evron 1991, 1998, and see below), bracketing the (Early) Natufian layer roughly between 16,000–12,400 cal BP. As part of the renewed excavation in the NE terrace, we launched a dating program in order to gain important insights into the duration and intensity of site occupation and its subphases, which will enable the testing of hypotheses pertaining to Natufian sedentism and the impact of abrupt environmental changes in the terminal Pleistocene. Provenance and Sampling During the renewed excavations at el-Wad terrace, a series of samples for 14C dating were collected from distinct proveniences covering a significant part of the attained Early Natufian sequence, mostly in the West Area of the excavation (Figure 1). In this part of the site, a thick Natufian sequence has been exposed, containing extensive Early Natufian stone architecture, overlain by Early Natufian levels with some living surfaces and stone installations, subsequently overlain by relatively thin Late/Final Natufian sediments with no built features. A group of Late/Final Natufian burials terminates the Natufian sequence at the NE terrace (Weinstein-Evron et al. 2007, in press; Yeshurun et al., in press). New 14C Dates for the Early Natufian of el-Wad Terrace 815 Figure 1 Map showing the location of el-Wad, a general plan of the site, and a plan of the renewed terrace excavation showing the Late Early Natufian (LEN) layer. The architectural complex of Wall I (large “terrace wall”) delimits numerous living levels, of which a portion of Structure II (Wall II) is shown in the East Area of the excavation and a rich living surface overlying it is depicted in the West Area. 816 M Weinstein-Evron et al. The samples (bone and charcoal) are derived from living floors, installations, and burials. Table 1 lists the provenience details for each sample (season of excavation, grid square and basket, “locus” if applicable, and elevations below the excavation datum). Note that as the layers dip NE (Figure 2), the elevation data can be safely correlated only within specific grid subsquares. Next, the table lists a short stratigraphic and contextual description of the samples. All of the samples come from the Natufian Unit 2, which is the main archaeological entity of the terrace. This layer, which correlates with Garrod’s layer B, contains a series of Natufian occupations exhibiting architectural features, graves, numerous finds, and virtually no signs of later intrusions. Unit 2 is provisionally divided into phases based on the intra-Natufian architectural stratigraphy. Phases W-0 through W-4 contain rich levels with a few built installations. Phase W-5 includes well-preserved occupation levels, while phases W-6 and W-7 display massive stone architecture including walls, numerous stony floors, and installations. Preliminary sorting of the lithics (Figure 2) points to the uppermost phase/s of Unit 2 and the bottom of Unit 1 as being of Late/Final Natufian age (dominated by short, abruptly retouched lunates and many microburins) and the rest of Unit 2 as belonging to the Early Natufian (dominance of Helwan-retouched lunates, abruptly retouched lunates being longer, and microburins are very rare). The excavation of the Natufian layers is still ongoing. We have attained the bottom of the Early Natufian layer in 2 small soundings, where only sterile earth and bedrock underlie it, similar to Garrod’s descriptions (Weinstein-Evron 2009; Weinstein-Evron et al., in press). Altogether, 46 samples (34 bones and 12 charcoals) were selected for dating, but only 11 (5 bones and 6 charcoals) were finally processed and yielded ages. A strict prescreening selection has been applied to determine the quality of the sample material (Eckmeier et al., these proceedings). Most 14C samples are derived from uninterrupted living surfaces, from the central (lines 6–7) and northern (lines 8–9) parts of the West Area of the excavation (Figure 1), from above, inside, and outside Structure II. A single sample was collected from a living surface in the East Area of the excavation. Additionally, 2 human burials were sampled: the femur of the newly unearthed Early Natufian Homo 132 (RTT-6114) and a charcoal piece (RTT-6116) from the burial of Homo 101 (a young male adult; Weinstein-Evron et al. 2007; Bachrach et al., in press). While the latter was attributed to a Late Natufian burial ground, its location within a cavity in the bedrock south of Wall I and its burial mode raise the possibility that it may belong to the Early Natufian (Weinstein-Evron et al. 2007). Figure 2 Section presenting the general stratigraphy of the terrace (marked as the N-O line in Figure 1). The lithic-based cultural division is depicted in the inset, showing the proportions of Helwan-retouched and abruptly retouched lunates, as well as microburins. New 14C Dates for the Early Natufian of el-Wad Terrace 817 Ta bl e 1 D es cr ip tio n of th e ar ch ae ol og ic al c on te xt a nd 14 C a ge s f or th e el -W ad te rr ac e sa m pl es . C al ib ra tio n us in g O xC al v 4. 15 (B ro nk R am se y et a l. 20 01 ; B ro nk R am se y 20 09 ) a nd th e In tC al 09 c al ib ra tio n cu rv e da ta (R ei m er e t a l. 20 09 ). La b # (R TT -) M at er ia l Ye ar Sq . B as . Lo cu s El ev at io n be lo w da tu m (c m ) U ni t p ha se le ve l, Ex ca va tio n ar ea R em ar ks A ge B P C al a ge B P (1  ) C al a ge B P