Two Early Epipaleolithic Units in Area C at Yutil al-Hasa in the Wadi al-Hasa, Western Highlands of Jordan

53

IntroductionYutil al-Hasa (WHS 784) is a collapsed rockshelter

situated in the main Wadi al-Hasa drainage in the Western Highlands of Jordan (Figure 1). It was first recorded duringMacDonald’s survey of the region in 1982 (MacDonald et al. 1983), and was recognized as an Epipaleolithic occupation by Gary Rollefson, who participated in that survey. It was described as a Geometric Kebaran “A” site in the terminology current in the early 1980s. As a result of its designation, the site was chosen for testing by the Wadi Hasa Paleolithic Project (WHPP) in 1984, although the two areas excavated (A and B) yielded late Upper Paleolithic artifacts rather than Epipaleolithic materials (Clark et al. 1987; 1988; Olszewski et al. 1990).

In 1993, further excavations by the WHPP at Yutil al-Hasa yielded Early Epipaleolithic (Areas C and E) and Late Epipaleolithic (Area D) occupations (Clark et al. 1994; Olszewski 1997; Olszewski et al. 1994), thus locating the occupation recognized by Rollefson. Following this, in 1998, the Eastern Hasa Late Pleistocene Project (EHLPP) added new units in Areas A, C, and D (Coinman et al. 1999). In 2010, the Western Highlands Early Epipaleolithic Project (WHEEP) returned to Yutil al-Hasa, excavating additional units in Area C and units in Area F, which also yielded Early Epipaleolithic materials (Olszewski and al-Nahar 2011). The author excavated at Yutil al-Hasa duringall these seasons and projects, and analyzed all the lithics discussed in this paper. In today’s terminology, the Early Epipaleolithic at Yutil al-Hasa is the Nebekian (Olszewski 2006). The geometrics recognized by Rollefson in 1982constitute a small proportion of the microlith component in Nebekian assemblages; they are mainly narrow trapezes with truncated ends formed by microburin scars.

Two Early Epipaleolithic Units in Area C at Yutil al-Hasa in the Wadi al-Hasa, Western Highlands of Jordan

Deborah I. Olszewski

The Early Epipaleolithic at Yutil al-Hasa (WHS 784) was recognized by Gary Rollefson during the MacDonald surveyin 1982. The site actually has three occupations (late Upper Paleolithic, Early and Late Epipaleolithic) that have beentested on four occasions by three different projects. This paper discusses two adjacent units which were dug in AreaC (Early Epipaleolithic), one in 1993, and one in 2010. Comparison highlights some differences due to excavationmethods in the two different seasons, while overall discussion examines the nature of the Early Epipaleolithic at Yutilal-Hasa in the context of the site and the region.

Area C: Units C and C2010-2One of the difficulties in excavating at Yutil al-Hasa

is the massive boulders across much of the surface of the site. These likely represent roof fall from a rockshelter inwhich the Early Epipaleolithic and late Upper Paleolithic occupations were situated, but may also represent roof collapse from an upper rockshelter (Area D) and/or simply rock fall from the bedrock ledge above the site. As a result, there are limited portions of the site into which excavation units can be placed, and even these exposed sections are not always large enough to accommodate more than one 1 × 1 m unit.

Over the course of three excavation seasons in 1993, 1998 and 2010, however, the WHPP, EHLPP and WHEEP were able to place several units within Area C. Of these, three were nearly contiguous (C from 1993, C98-1 from 1998, and C2010-2 from 2010), although of differentsizes (Figure 2). All are situated at what would have been the back of the rockshelter. Units C and C98-1 were both 1 × 1 m, while C2010–2 was 50 × 50 cm. Here, the lithic assemblages from the upper 50 cm in the adjacent units C and C2010–2 are used. These were chosen because theyshare the same stratigraphy (the natural layers in unit C98-1 differ somewhat).

Excavation protocols in 1993 differed from thoseemployed in 2010. Unit C was dug using arbitrary 10 cm levels (unless a new natural layer was encountered before 10 cm was reached) across the entire unit, while unit C2010-2 was excavated in 3 cm arbitrary levels. In 2010, the strategy was to dig a 1 × 1 m unit using 3 cm arbitrary levels in each 50 × 50 cm quad of the unit. However, as unit C2010-2 was only 50 × 50 cm in size, it served as a ‘quad.’ One factor with potential to affect lithic recovery,

54 Settlement, Survey, and Stone

Figure 1. Looking north at Yutil al-Hasa in 2010: (a) close-up of Area C; (b) overview of the site from the wadi bottom, with Area C indicated by the arrow (photos by D. I. Olszewski)

Olszewski: Two Early Epipaleolithic Units at Yutil al-Hasa in the Wadi al-Hasa 55

Figure 2. Plan view of the Area C units from the excavations in 1993, 1998, and 2010. Note that due to the boulders, placement of units often incorporated parts of several grid squares (illustration by D. I. Olszewski)


C (1 x 1 m) C2010-2 (50 x 50 cm)

% n % n

Tools

Endscraper 1.9 7 2.2 4

Burin 1.9 7 2.8 5

Perforator - - - -

Backed piece - - 1.1 2

Truncation 0.6 2 1.1 2

Notch/denticulate 6.2 22 6.1 11

Retouched piece 10.4 37 4.4 8

Nongeometric microlith 62.3 221 67.2 121

Geometric microlith 1.4 5 1.7 3

Microlith fragment 15.2 54 13.2 24

Special tool - - - -

Tools total 100 335 100 180

Cores

Flake 14.8 4 8.7 2

Blade/bladelet 37 10 60.9 14

Mixed 22.2 16 21.7 5

Fragment 26 7 8.7 2

Cores total 100 27 100 23

Debitage

Flake 12.6 615 6.2 195

Blade 3.2 157 2.9 90

Bladelet 21.5 1050 10.4 325

Small flake 44.2 2154 65.3 2039

Burin spall 0.3 16 0.9 29

Microburin 2.9 140 6.1 189

Shatter 15.3 746 8.1 254

Debitage Total 100 4878 100 3121

Grand Total 5260 3324

particularly of small pieces, was screen size. In 1993, screens of about 5–6 mm mesh were used, while in 2010, the screen mesh was 2 mm. The 2010 WHEEP season also used a totalstation to point-provenence all materials larger than about 2.5 cm, although the results of piece plotting are not the subject of this paper.

Radiocarbon dates were obtained for the late Upper Paleolithic and Late Epipaleolithic occupations at Yutil al-Hasa. Charcoal samples from the 2010 WHEEP excavations in Area C have been submitted for radiocarbon dating; results are pending. The types of nongeometric microliths in theArea C units (see below) are those reflecting the Nebekian


Early Epipaleolithic, which elsewhere has been radiocarbon dated to the interval between about 25,000 to 17,400 cal BP (Byrd and Garrard 2013, 356–65, Table 9.3; Olszewski 2003, 232, Table 19.1; Richter et al. 2013).

As noted above, the lithics discussed here are from the upper 50 cm of deposits in units C and C2010-2. A basic breakdown of the tools, cores, and debitage is shown in Table 1. As the units differ in size, it is the frequencies ofthe artifacts that are most pertinent with respect to assessing the materials recovered. Overall, the tool classes are virtually mirrors of each other for the two units. The one exceptionis the slightly higher frequency of nongeometric microliths in C2010-2. This is undoubtedly due to the smaller meshscreens used in 2010, which would have facilitated the recovery of elements such as nongeometric microliths. What is intriguing in this respect is that there is not a higher frequency of microlith fragments from C2010-2, which might be expected given the screen mesh size. Within the nongeometric microlith component, the most frequent type is exceptionally narrow forms called ‘attenuated curved’ by the author (‘double arched’ in other terminologies), which constitute 42 % to 44 % of each unit’s nongeometric microliths. Rare types include La Mouillah points, Ouchtata bladelets, Dufour bladelets, and Qalkhan points; most of these are present as fewer than six artifacts combined totals from both units. The small number of geometrics includesa type defined by the author as a microburin trapeze; thisrefers to the fact that the ends of the piece are microburin scars which are angled. In all likelihood, these types of geometrics are simply a variation on the attenuated curved nongeometrics (which have microburin scars that are curved rather than angled).

In the core component of the assemblage, there are interesting differences between the two units. This is seenprincipally in the much higher frequency of blade/bladelet cores in C2010-2, and it is of note that the 50 × 50 cm unit C2010-2 yielded almost as many cores in absolute count as did the 1 × 1 m unit C. Most cores are single platform (55 %–71 %), with the most common removal being bladelets.

Despite the presence of more blade/bladelet cores in C2010-2, bladelet debitage is more frequent in C (Table 1). There also are more flakes and shatter in C. Two othernoticeable differences between the adjacent units are mostlikely due to recovery biases related to screen mesh size. The smaller mesh used in 2010 appears to have resultedin the recovery of exceptionally high amounts of small flakes (≤ 25 mm in size) and microburins. Nearly all themicroburins are regular (true); there are also a small number of Krukowski and piquant trihèdre types. The extremelyhigh incidence of microburins in C2010-2 also is reflectedin the Imbtr (restricted microburin index), which is 60.4 in C2010-2 and 38.3 in C. The Imbtr excludes microlithfragments from the calculation. Although the indices are

substantially different, both reflect a heavy emphasis on theuse of microburin technique to segment bladelets for the manufacture of microliths.

DiscussionComparison of Unit C and C2010-2

One factor often discussed by archaeologists is the representativeness of the artifact samples obtained from excavation units. These include issues potentially relatedto activity patterns in areas of a site that result in differentfrequencies of artifact types, as well as taphonomic processes that can winnow out (or in) certain items, particularly small elements that are easily transported by surface water runoffor through the burrowing activities of rodents and insects. In the case of units excavated by different projects at differenttimes, there are also potential issues that can arise due to the methodologies used.

At Yutil al-Hasa, the Area C excavations of units C and C2010-2 were immediately adjacent contexts, which likely means that any differing frequencies of artifacts in thetwo units is not due to the accumulation of the remains of different activities. Moreover, during both field seasons,disturbances due to rodents were not observed in either unit, suggesting that this particular type of taphonomic process was not in play. The differences observed must, therefore,be due to the effect of smaller mesh screens used in 2010and/or winnowing of small lithic elements.

The winnowing taphonomic process is possiblyresponsible for the high incidence of microburins in unit C2010-2 because the northern portion of the eastern profileof unit C was described as containing a colluvial pocket of materials (J. Schuldenrein, personal communication 1993), likely meaning that small elements such as microburins were transported as part of this colluvial event. This eastern profileis the western wall of unit C2010-2 and during excavations in this unit, this pocket of materials was observed in the northern portion of the unit and excavated as a separate context. When the Imbtr is calculated for the colluvial pocket context versus the other natural layers in the upper 50 cm of unit C2010-2, the indices are 64.4 for the colluvial pocket and 51.5 for the other layers. The role of taphonomyin C2010-2 also can be seen using a comparison to another of the units excavated in 2010 in Area C (C2010-1), which is situated about 2 m to the southwest and for which the smaller mesh screens were also used. The Imbtr for the upperdeposits in C2010-1 is 42.2, which is more comparable to unit C than to unit C2010-2.

Although there are some differences in the frequenciesof lithic classes/types in the comparison between units C and C2010-2, most of these do not appear to be significantdespite the fact that recovery techniques used in unit C (1993) were not as fine-grained as those used 17 yearslater in unit C2010-2. Even in the context of the role of


taphonomic processes, seen most clearly in unit C2010-2, the aggregated materials do not impact the attribution of the units to the Nebekian Early Epipaleolithic. What these taphonomically aggregated materials do influence,however, is any inclination to interpret them as discrete anthropomorphic activity distribution patterns at the site.

Interpretation of Area C at Yutil al-Hasa and the regional context

The five units excavated in Area C at Yutil al-Hasaduring the 1993, 1998, and 2010 field seasons all yieldedlithic assemblages that reflect occupations during theEarly Epipaleolithic. These assemblages are dominated bynongeometric microlith types (especially attenuated curved bladelets) made using the microburin technique, with much rarer examples of larger tools such as endscrapers, burins, truncations, and so forth. As an aggregate, they are best referred to as Nebekian and thus early in the sequence of the Early Epipaleolithic (see discussion of the Nebekian in Byrd and Garrard 2013, 374–80).

The density of lithic artifacts by weight in Area C atYutil al-Hasa is not particularly high (c. 5.6 kg for units C and C2010–2 together). In combination with the high frequency of microliths and a relatively low incidence of cores, this may suggest that occupational episodes at the site were sporadic and/or short-term. The emphasis on microlithspossibly reflects specialized tool kits geared toward huntingactivities (as suggested for the Nebekian in the Azraq Basin region, see Byrd and Garrard 2013, 389), although such an interpretation would need to be substantiated by other types of analyses, e.g., residue or microwear studies. Faunal materials from Area C at Yutil al-Hasa were not particularly well preserved, but did include gazelle, tortoise, aurochs, equids, and wild goat (listed in order of decreasing NISP).

There are several sites in the Wadi al-Hasa regionwhich have lithic materials suggesting that they are Early Epipaleolithic. These include Tor at-Tareeq in the easternbasin of the Wadi al-Hasa, about 3 km to the SSE of Yutil al-Hasa (Clark et al. 1987, 52–67; 1988, 253–72; Neeley et al. 1998; Olszewski et al. 2000), Tor Sageer in a tributary wadi to the Wadi al-Hasa, about 2 km to the NNE of Yutil al-Hasa (Coinman et al. 1999, 16–19; Olszewski et al.1998, 59–61), and KPS-75 on the Kerak Plateau, roughly 10 km to the NNE of Yutil al-Hasa (al-Nahar et al. 2009). Based on the frequencies of the attenuated curved microliths at these sites, Area C (including all excavated units) at Yutil al-Hasa compares most favorably with the lower occupation at Tor at-Tareeq, which has a series of radiocarbon dates ranging from about 20,500 to 18,700 cal BP (Neeley et al. 1998). This would place Yutil al-Hasa Area C somewhat later than

the Tor Sageer occupations, which date from about 25,500 to 24,200 cal BP (Olszewski 2003, 232); the early dates for Tor Sageer suggest that it can be described as what Byrd and Garrard (2013, 369) call the Initial Epipaleolithic. Although KPS-75 has not been dated, both the lower and upper occupations at the site have quite a few Qalkhan points (4.9 % and 3.1 %, respectively), perhaps indicating that this site should be designated as Qalkhan Early Epipaleolithic. Given that recent work in the Azraq region shows that Qalkhan sites are chronologically later than the Nebekian (Byrd and Garrard 2013, 372), this suggests that Yutil al-Hasa Area C is earlier than both the occupations at KPS-75.

Perhaps one of the most interesting features of the Early Epipaleolithic (Initial, Nebekian and Qalkhan) in the Wadi al-Hasa region is its relative rarity in a landscape that had paludal contexts (springs, marshes, and ponds) that one might expect to be a major attractor in an otherwise cool and dry late Pleistocene ecology. The fact that the EarlyEpipaleolithic deposits at each of the sites (Yutil al-Hasa Area C, Tor Sageer, Tor at-Tareeq, and KPS-75) range from c. 80 cm to more than a meter thick does suggest that these were favored locales to which people returned quite frequently over what was probably a considerable period of time at each site (that is, over at least several generations, if not longer). There may have been, of course, sites thathave disappeared due to erosion in the wadi system since the late Pleistocene, particularly if such sites were situated in the marls present in the eastern portion of the basin. Other factors relevant to the paucity of Early Epipaleolithic sites may be the poor visibility of diagnostic microliths during surveys, particularly for ephemeral occupations – and thus a lack of identification of sites belonging to this chronologicalperiod – as well as the possibility that the landscape was very lightly populated so that prehistoric groups in this region are largely invisible except at favored locales such as Yutil al-Hasa Area C.

AcknowledgementsFunding for the excavations at Yutil al-Hasa was

provided by grants from the National Science Foundation to G. A. Clark for the 1984 and 1993 excavations, to N. R. Coinman and D. I. Olszewski for the 1997 season, and to D. I. Olszewski and M. al-Nahar for the 2010 excavations. Specialist analyses for the 1997 season were funded by grants to D. I. Olszewski from the National Geographic Society and the Wenner-Gren Foundation. Logistical support for the 2010 season was provided by the Institute of Archaeology at Jordan University, by Hashemite University, and by the American Center for Oriental Research in Amman. This isEHLPP Contribution #35 and WHEEP Contribution #4.


ReferencesByrd, B. F. and A. N. Garrard (2013) Region patterns

in Late Palaeolithic chipped stone production and technology in the Levant. In A. N. Garrard and B. F. Byrd (eds), Beyond the Fertile Crescent. Late Palaeolithic and Neolithic Communities of the Jordanian Steppe. TheAzraq Basin Project. Volume 1, Project Background and the Late Palaeolithic (Geological Context and Technology), pp. 350–93. Levant Supplementary Series Volume 13. Oaksville, CT: Oxbow and CBRL.

Clark, G. A., Lindley, J. H., Donaldson, M., Garrard, A., Coinman, N., Schuldenrein, J., Fish, S. and Olszewski, D. I. (1987) Paleolithic archaeology in the southern Levant. A preliminary report of excavations at Middle, Upper and Epipaleolithic sites in Wadi el-Hasa, west-central Jordan. ADAJ 31, 19–78; Plate, 547.

— (1988) Excavation at Middle, Upper and Epipaleolithic sites in Wadi Hasa, west-central Jordan. In A. N. Garrard and H. G. Gebel (eds), The Prehistory of Jordan. The Stateof Research in 1986, pp. 209–85. BAR International Series 396(i). Oxford: Archeopress.

Clark, G. A., Olszewski, D. I., Schuldenrein, J., Rida, N. and Eighmey, J. (1994) Survey and excavation in the Wadi al-Hasa: A preliminary report of the 1993 fieldseason. ADAJ 38, 41–55.

Coinman, N. R., Olszewski, D. I., Abdo, K., Clausen, T., Cooper, J., Fox, J., al-Nahar, M., Richey, E. and Saele, L. (1999) The Eastern Hasa Late Pleistocene Project:Preliminary report on the 1998 season. ADAJ 43, 9–25.

MacDonald, B., Rollefson, G. O., Banning, E. B., Byrd, B. F., and D’Annibale, C. (1983) The Wadi al-Hasaarchaeological survey 1982: A preliminary report. ADAJ 27, 311–23.

al-Nahar, M., Olszewski, D. I. and Cooper, J. B. (2009) The2009 excavations at the Early Epipaleolithic site of KPS-75, Kerak Plateau. Neo-Lithics 2/09, 9–12.

Neeley, M. P., Peterson, J. D., Clark, G. A., Fish, S. K. and Glass, M. (1998) Investigations at Tor al-Tareeq: An Epipaleolithic site in the Wadi el-Hasa, Jordan. Journal of Field Archaeology 25/3, 295–317.

Olszewski, D. I. (1997) From the late Ahmarian to the early Natufian. A summary of hunter-gatherer activities atYutil al-Hasa, west-central Jordan. In H. G. K. Gebel, Z. Kafafi and G. O. Rollefson (eds), The Prehistory ofJordan II: Perspectives from 1997, pp.171–82. SENEPSE 4. Berlin: ex oriente.

— (2003) The conundrum of the Levantine Late UpperPalaeolithic and Early Epipalaeolithic: perspectives from the Wadi al-Hasa, Jordan. In N. Goring-Morris and A. Belfer-Cohen (eds), More Than Meets the Eye:Studies in Upper Palaeolithic Diversity in the Near East, pp. 230–41. Oxford: Oxbow Books.

— (2006) Issues in the Epipaleolithic: the Madamaghan, Nebekian, and Qalkhan (Levant Epipaleolithic). Paléorient 32/1, 19–26.

Olszewski, D. I., Clark, G. A. and Fish, S. K. (1990) WHS 784 X (Yutil al-Hasa): A late Ahmarian site in the Wadi Hasa, west-central Jordan. Proceedings of the Prehistoric Society 56, 33–49.

Olszewski, D. I., Coinman, N. R., Schuldenrein, J., Clausen, T., Cooper, J., Fox, J., Hill, J. B., al-Nahar, M. and Williams, J. (1998) The Eastern Hasa Late PleistoceneProject: Preliminary report on the 1997 season. ADAJ 42, 53-74.

Olszewski, D. I., Cooper, J. B., Jansson, H. and Schurmans, U. (2000) A third season of excavation at Tor al-Tareeq (WHS 1065), an Early and Middle Epipaleolithic site in the Wadi al-Hasa, Jordan. Neo-Lithics 2–3/00, 14–16.

Olszewski, D. I. and al-Nahar, M. (2011) A fourth season at Yutil al-Hasa (WHS 784): Renewed Early Epipaleolithic excavations. Neo-Lithics 1/11, 30–4.

Olszewski, D. I. Stevens, M. N., Glass, M., Beck, R. F., Cooper, J. and Clark, G. A. (1994) The 1993 excavationsat Yutil al-Hasa (WHS 784): An Upper/Epipaleolithic site in west-central Jordan. Paléorient 20/2, 129–141.

Richter, T., Maher, L., Garrard, A. N., Edinborough, K., Jones, M. and Stock, J. (2013) Epipalaeolithic settlement dynamics in southwest Asia: New radiocarbon evidence from the Azraq Basin. Journal of Quaternary Science 28, 467–79.

Two Early Epipaleolithic Units in Area C at Yutil al-Hasa in the Wadi al-Hasa, Western Highlands of Jordan

Documents