In July–August 2013, a trial excavation was conducted near the abandoned village of Khirbat ‘Asafna (East; Permit No. A-6848; map ref. NIG 209622–722/736096–293; Fig. 1), after archaeological remains were identified while deepening a defunct fish pond for renewed use. The excavation, on. behalf of the Israel Antiquities Authority and financed by the Cross-Israel Highway Company Ltd., was directed by E.C.M. van den Brink, with the assistance of D. Kirzner and A. Yaroshevich (field supervisors of Areas A and B respectively), Y. Amrani and E. Bachar (administration), R. Mishayev, M. Kahan, A. Hadjian and M. Kunin (surveyors), A. Peretz (field photography), N. Marom (archaeozoology) and H.K. Mienis (shells).
The excavation uncovered poorly preserved remains from the Middle Bronze Age II, a living surface from the Early Chalcolithic period (Wadi Rabah culture), a settlement dating from the Pottery Neolithic period and flint items of the early Pre-Pottery Neolithic B. The excavation was located in a natural depression that in 1952–1953 was enclosed with an embankment (c. 3 m high, top at 17.5 m asl) to create a large fish pond (100 × 150 m; Fig. 2). The pond was near Kibbutz Sha‘ar Ha-‘Amaqim, about 300 m east of the present streambed of the perennial Nahal Qishon and Road 70, which today runs alongside it. The pond remained in operation until the 1990s.
Although the area was surveyed several times (Na‘or 1961
; Olami, Sender and Oren 2004
; Permit Nos. A-3989, A-3996), the site, which was sealed under a layer of alluvial clay sediments (thickness 0.4–2.0 m), remained unrecognized until A. Sa‘id (IAA internal report, 22/04/2013) identified archeological remains in eight out of 59 test trenches mechanically excavated within the pond. Less than 200 m to the southeast lies the site of Elro’i, where surface finds from the Middle Paleolithic and Neolithic periods were recorded in the past (Olami, Sender and Oren 2004
:49*, Site 127).
On the basis of this information, a layer (thickness c. 0.4 m) of compact, archaeologically sterile, alluvial clay sediments, which was deposited over the years by Nahal Qishon, was mechanically removed in the extreme southeast part of the former pond (c. 20 × 40 m; Area A), down to 15.45–15.15 m asl. A similar topsoil layer (thickness 1.5–2.0 m) was removed prior to excavation in the southwest part of the pool (Area B), 80 m north of Area A, down to c. 13.70 m asl. Area A comprised 16 squares (4 × 4 m each; 91 loci in total), and Area B—four adjoining squares (5 × 5 m each; 24 loci in total).
Early Pre-pottery Neolithic B and the Pottery Neolithic periods
Building remains were unearthed in Area A (Fig. 3), a few centimeters below the surface (15.15–15.45 m asl from north to south): a number of dense spreads of small, natural, mainly angular stones and several medium- and large-sized fieldstones, all of limestone. A few of the small stone-spreads were beddings for floors of (storage?) installations and of living surfaces. These were attached to, or associated with, rectangular structures; most of the medium- and large-sized fieldstones turned out to be part of these structures. The remains belonged to at least four distinct rectangular architectural units that formed part of one, possibly two, so-called Pottery Neolithic, ‘Yarmukian’ courtyard buildings (Fig. 4), a type best known from the PN site at Sha‘ar Ha-Golan.
Only two rooms were fully exposed, while three additional wall segments indicated at least another two to three rooms that remain unexcavated. In a corner of one of the excavated rooms (Fig. 5), a large storage jar from the Pottery Neolithic period (c. 5800 BCE) was found in situ. It lay under the burnt debris of mud-brick walls that once rested on the stone foundations and caved in, possibly due to an accidental fire. A small, but notable difference between the architecture of Sha‘ar Ha-‘Amaqim and Sha‘ar Ha-Golan is the use of flat, square mud-bricks in the former and plano-convex mud-bricks in the latter. A square installation, possibly used for storage, was attached to the southwest corner of the room on the outside. Clearly circumscribed patches of small stones nearby indicate circular installations that were probably located in the internal courtyard of the house. An earlier phase containing considerable architectural remains was identified underneath the two rooms and some of the isolated stone-wall foundations, but time did not allow its exposure.
Besides pottery dating from the Pottery Neolithic period, the finds included a variety of flint tools, ground stone tools—mainly grinding and pounding stones, as well as pestles—and animal bones belonging to cows, pigs and gazelles. This combination indicates processing of agricultural produce as well as husbandry and hunting. The finds from Area A included various flint items that could be dated with confidence to the early Pre-Pottery Neolithic B period (c. 6000 BCE). They included several arrow heads (Fig. 6), including at least two of an early type known as Helwan points, tranchet axes and sickle blades on blanks obtained through a bipolar reduction technique. Early PPNB sites are very rare in Israel, and the finds are therefore a welcome addition to the sparse data.
Early Chalcolithic (Wadi Rabah) Period
This period was represented in all four squares of Area B. However, due to the relatively high water table in the area, manual excavation of the southern part of Area B was impossible, and excavations were limited to its northern part. The remains comprise a layer of small stones, considerably inclined to the northwest—possibly a pavement or a living surface. The original edge of this layer was identified in three squares, making it possible to suggest that it was oval and oriented northeast–southwest. Since the northeastern edge was not exposed, this reconstruction remains tentative. One possible explanation for the sharp inclination of the paving is that this area was once a terrace of Nahal Qishon, although today the river flows c. 300 m to the west.
In three squares, concentrations of medium–large limestone fieldstones were found interspersed within the layer of small stones and above it. The function of these patches remains unclear. Two of the large fieldstones bear a cupmark (diam. c. 8 cm). Some of the large stones embedded into the pavement may have been used as flint knapping surfaces, as the large flint assemblage found in this layer clearly indicates on-site flint knapping. Dark clay sediment containing a large number of sweet-water shells covered the pavement.
Only a few pottery fragments dating from the Early Chalcolithic period were recovered from the two areas, but the associated flint assemblage is sizeable and rather homogeneous. The numerous of cores, core-trimming elements, primary elements and other debitage groups indicate on-site knapping activity. The assemblage is dominated by ad-hoc tools, mainly massive scrapers and denticulated items. Among the diagnostic tools are bifacials (axes and adzes), tabular scrapers and sickle blades, mostly denticulated, backed and truncated items (Fig. 7). Three polished axes, one of them made on tabular flint, constitute an interesting group within the flint tool assemblage. The majority of the flint artefacts are fresh; a few were patinated and slightly abraded.
Middle Bronze Age II
Dispersed remains appeared only in the eastern part of Area B, and included a few concentrations of medium–large limestone fieldstones. The function of these poorly preserved remains is not clear; their dating to the MB II is based on a scant amount of diagnostic potsherds, mainly rims of storage jars. The associated flint artefacts are few, but include a large sickle blade with a curved back, made on a blank resembling the Canaanean blade industry. The remains were covered by a layer of dark clay containing large amounts of small shells characteristic of a sweet-water environment, indicating an episode of flooding by the nearby Nahal Qishon.
The presence of MB II remains at Khirbat ‘Asafna is not surprising, since several rock-cut tombs of this very period were excavated on the southern slope of the nearby Tel Me‘ammer (Tell el-‘Amar; Olami, Sender and Oren 2004
: Site 73), a few kilometers to the north (Drucks 1982
The remains unearthed in the excavation belonged predominantly to three periods: the Pottery Neolithic, the Early Chalcolithic period (the Wadi Rabah culture) and the Middle Bronze Age II. The remains of the Pottery Neolithic period include dwelling remains—four units, probably rooms of a Yarmukian courtyard-type building or buildings—with several installations that were located in Area A. The variety of ground stone items, flint tools and animal bones on site indicates a mixed subsistence economy of agriculture and husbandry, with some hunting. The availability of fresh water from the nearby spring (Ma‘ayan Ha-Zor‘im; ‘Ein el-Jafar; I. Na’or, Director of the Archaeological Museum at Kibbutz Sha‘ar Ha-‘Amaqim, pers. comm.) and from Nahal Qishon undoubtedly facilitated settlement in this location. The remains of the Early Chalcolithic period comprise a surface built on an ancient terrace that overlooked Nahal Qishon (Area B), where flint knapping activities took place. Sparse Middle Bronze Age II were restricted to Area B.
Although time constraints allowed hardly any probing, it is clear that there is an earlier stratum with architectural remains beneath the Area A structures thus far exposed. The sporadic finds of early Pre-Pottery Neolithic B flint tools, which included several arrowheads, indicate human activities in this area during the early PPNB.
Drucks A. 1982. Early Tombs at Tel ‘Amr. ‘Atiqot
) 8:1–6 (English summary, p. 1*).
Na’or Y. 1961. The Spring of Sha‘ar Ha-‘Amaqim: A Survey. JIPS 2:23–26 (Hebrew).
Olami Y., Sender S. and Oren E. 2004. Map of Yagur (27) (Archaeological Survey of Israel). Jerusalem.