In August 2006 a trial excavation was conducted in the western part of Qibbuz Sha‘alvim (Permit No. A-4879*; map ref. NIG 198117–397/641672–2146, OIG 148117–397/141672–2146), in an area slated for expansion. Ancient agricultural installations were revealed in a previous survey and soundings. The excavation, on behalf of the Israel Antiquities Authority and with the financial backing of the qibbuz, was directed by Y. Arbel, with the participation of L. Talmi (area supervision), E. Bachar (administration), A. Hajian (surveying), H. Khalaily (lithics), L. Zeiger (flint drawing), M. Shuiskaya (pottery drawing) and P. Gendelman (pottery reading).
A number of excavations were conducted at Sha‘alvim in the past. Among the antiquities uncovered was a Samaritan synagogue with mosaic inscriptions that dates to the fourth–fifth centuries CE (Louis M. Rabinowitz Fund for the Exploration of Ancient Synagogues Bulletin
I [Jerusalem 1949], 26–30). Two Middle Bronze Age II burial caves were excavated yielding numerous pottery vessels, metal, stone and alabaster artifacts (HA
22:9–14). Two additional burial caves from the same period were exposed along the eastern side of the qibbuz together with building remains and installations from the Late Bronze Age and the Iron Age II (HA-ESI 120
). In addition, an animal pen, stone clearance heaps and cisterns were excavated in the area (HA-ESI 119
In the current excavation the installations located in the survey were cleaned; other installations were identified during the excavation (Fig. 1). These included olive-oil extraction installations (bodedot), winepresses, cupmarks, a pit, a watchtower and stone clearance heaps. Seventeen locations were excavated for the purpose of locating installations below the alluvium. A survey was also conducted in which flint implements were collected; two trial squares were excavated in order to find the production zone for the flint tools.
Five olive-oil extraction installations (bodedot) of different sizes were exposed that were hewn in bedrock outcrops. These consisted of a shallow, elliptical press bed (depth 10 cm) and a round and relatively deep cupmark which were linked by a channel. In some of the collecting basins a small shallow settling depression was found. The finds recovered from inside them are associated with the alluvium and not with the activity of the installations themselves. The bodedot were located in clusters along the same contour line, probably near the cultivation plots or the settlement. The only exception was the installation at Point 11, which had a different shape and location and was probably used for another activity.
Point 4: a pressing surface (1.05 × 1.05 m), collecting basin (diam. 0.55 m, depth 0.30 m) and a settling niche (diam. 0.15 m, depth 0.20 m; Fig. 2).
Point 5: a pressing surface (0.60 × 0.80 m), collecting basin (diam. 0.50 m, depth 0.20 m) and a settling niche (diam. 0.13 m, depth 0.10 m). The bodeda was hewn in a large bedrock mass that was moved from its original place under unclear circumstances and is currently located above a cistern (denoted by the same point number).
Point 7: a pressing surface (0.90 × 1.05 m), collecting basin (diam. 0.70 m, depth 0.30 m) and a settling niche (diam. 0.15 m, depth 0.20 m). A cupmark (diam. 0.45 m, depth 0.25 m) was found next to the collecting basin; the functional relationship between them is unclear.
Point 10: a pressing surface (0.80 × 1.10 m) and collecting basin (diam. 0.55 m, depth 0.30 m).
Point 11. The installation was irregular in shape and consisted of a pressing surface (0.60 × 1.30 m), elliptical collecting basin (0.45 × 1 00 m, depth 0.23 m) and a niche (diam. 0.15 m, depth 0.10 m; Fig. 3).
Two winepresses were exposed in the northern part where the slopes are steeper, while others were identified farther along the northern slope and were not excavated. The two winepresses included treading floors and rectangular collecting vats, without settling basins or other adjacent installations. The northern winepress (Point 1) had a treading floor (2.20 × 2.20 × 0.40 m; Fig. 4) and a collecting vat (0.70 × 1.20 × 1.00 m) that was used in another phase as a tomb as evidenced by its interior structure and a hewn rectangular frame for a cover that was approximately 10 cm thick. The collecting vat was found partly filled with alluvium that contained a few potsherds which postdate its use-phase.
The second winepress (Point 16) had a relatively wide treading floor (length of sides c. 2.5 m), an irregular shape and shallow depth (c.10 cm). Its style resembles the pressing surfaces of bodedot even though it was wider. The collecting vat was rectangular (0.95 × 1.80 × 0.45 m); its floor was not level and the quarrying of it, like that of the rest of the winepress, was never completed. A round rock-hewn pit (diam. 1.2 m, depth 0.70 m) that was probably used for storage was found nearby.
Three pairs of cupmarks were found (Points 3.1, 3.2 and 17; diam. c. 0.50 m, depth c. 0.30 m). A pair of cupmarks (Point 3.2) was enclosed by a shallow frame (length per side 0.70 m, width 0.10 m, depth 0.08 m; Fig. 5). A frame that was less well-preserved and slightly broader survived around another pair of cupmarks (Point 3.1). It seems that the frames were used to secure a wooden installation associated with the production or pressing activity or as a cover to protect some sort of product. Three other cupmarks of similar dimensions were found in the area. Two of them were next to a bodeda (Point 7) and one was near a winepress (Point 1). Small isolated cupmarks were found near a cistern (Point 5), a winepress (Point 16) and another bodeda (Point 4).
The pit was located in the upper central part of the area (Point 5) and below a bodeda that was hewn in bedrock. It was bell-shaped and consisted of a shaft that was excavated in the hard nari (diam. 0.8 m) that gradually became wider in the soft chalky layers. Found filled with soil, it was surveyed but not excavated. There were no remains of plaster on its exposed walls. A drainage hole for channels discovered on the surface was located next its eastern wall. The date of the pit and the relationship between it and the rest of the installations in the area are unclear.
Watchtower and Stone Clearance Heaps
A round structure (diam. c. 3.5 m) was located in the southern part of the site (Point 8) alongside a concentration of stone clearance heaps (Points 9, 13). The walls of the structure were built of large stones, some of which were dressed. During the course of the excavation half of the structure was removed and the core of its walls was revealed which included field stones in the soil, without any datable finds.
Two stone clearance heaps (Points 14. 15) and smaller concentrations of stone in the vicinity were surveyed. The stone clearance heaps were cleaned and removed but no datable finds were discovered. They were probably modern, related to activities conducted by the Jewish National Fund in the region. Next to the heaps and inside them were large ashlar stones (0.60 × 0.65 × 0.20 m) whose original provenance was unclear.
Remains of a quarry were identified (height c. 2 m; Point 2) on the northern slope of the site, near a winepress (Point 1); no datable finds were discovered in it. Other signs of rock cuttings were observed farther along the slope, outside the boundaries of the site.
Large quantities of flint, some of which were knapped, were discovered throughout the area. A trial excavation conducted where a relatively dense concentration of tools was observed (Point 6) did not reveal any clear-cut activity horizons. A flint collecting survey was conducted throughout the entire site in which 22 tools, 24 blades and over 100 flakes were collected. The flint implements were produced from the Mishesh Formation, characteristic of the Modi‘in region. The knapping was probably done at the site itself as evidenced by the large amount of debitage found there. The finds include five cores (Fig. 6:1), one of which bearing scars of flakes and blades that were knapped from it. The flint industry was primarily meant for the production of large flakes for use as everyday tools. Among the flint items are five sickle blades and sickle elements (Fig. 6:2–5), six retouched flakes, two scrapers (Fig. 6:6) and four burins (Fig. 6:7). Among the blades is a single Canaanean blade made of Eocene flint (Fig. 6:2) which was probably imported to the site. Three of the other flakes are geometric flakes c. 3 cm wide and more than 1 cm thick. The relatively crude sickle blades probably date to the Iron Age or even later.
Potsherds from different periods were found on the surface and next to the installations. At best they provide a general date for activity at the site. Among the potsherds identified are jars from the Persian or Hellenistic periods (Fig. 7:1–2), a pared lamp from the first century CE (Fig. 7:3), many fragments of ribbed jars from the Byzantine period, a glazed bowl (Fig. 7:4) and a jug handle (Fig. 7:5) from the Early Islamic period, and jar fragments from the Ottoman period (Fig. 7:6). Some of the body sherds give the impression that they date to even earlier periods.
The flint tools and pottery are consistent with the settlement picture at the site and its surroundings in the Bronze and Iron Ages as revealed in prior excavations, while the style of the winepresses and the bodedot indicate a possible date in the Roman and Byzantine periods. It should be noted that all of the bodedot were hewn along the same contour line on the high parts of the spur whereas the winepresses all occur on the northern part of the site where the slope is steeper. If we indeed assume that the installations operated simultaneously, then it is possible to understand the functional division of the installations in parallel to the deployment of the growing of olives and grapes.