During February–March and June–July 2006, renewed trial and salvage excavations were conducted between segments of Roads 8 and 12, opposite Horbat Hadat–Be’erit, within the Buchman compound in Modi‘in (Permit No. A-4735; map ref. NIG 2010/6420; OIG 1510/1420), in the wake of construction. The excavation, on behalf of the Antiquities Authority and financed by the Ministry of Housing and Construction, was directed by E.C.M. van den Brink (photography), with the assistance of U. ‘Ad (supervision Area C2), M. Mulokandov (supervision Area C3), Y. Ochayon and E. Bachar (administration), V. Essman, A. Hajan, V. Pirsky, T. Kornfeld, (surveying), T. Sagiv (field photography), Sky Balloons (aerial photography), N. Zak and I. Berin (drafting), S. Lev-Yadun (olive stones), P. Gandelman (Roman–Byzantine pottery), R. Avner (depictions in the miqve), J. Neguer, N. Davidoff and R. Abu-Dihab (preservation), M. Sa‘id (metal detector) and Y. Elisha and T. Lakichevic (supervision of heavy machinery).
Five 4 × 4 m probes were opened on the lower slopes of the hill and excavated down to bedrock. Several stone-built wall segments of unknown date and function were uncovered. Two of these were originally thought to be ancient terrace walls, but turned out to be segments of two parallel alignments of large boulders that bordered an ancient, east-northeast–west-southwest oriented road. A bell-shaped cave higher up the same slope could only be cursorily explored, yielding no clues as to its date or function.
Four sub-areas, labeled Areas C1–C4 were opened.
After the mechanical removal of a Hellenistic–Roman (Hashmonean) farmhouse, whose central and northern parts had been excavated by A. Onn and the late G. Parnos in 2004 (Permit No. 4069), remains of three karstic caves, c. 20 m from each other, were visible in the west section of the hillside, after it was cut by a bulldozer in preparation for Road 8. The caves are briefly described from north to south.
Cave 1. Only the very rear end of this cave (length c. 3 m, width 0.5 m, height 1.5 m) had survived. The few potsherds collected from the remaining fill dated its usage to the Late Chalcolithic period.
Cave 2 (Fig. 1). It is estimated that about one third of this cave was removed unnoticed, including the original entrance. Three layers were identified in the soil fill that was in the remaining cave area (length 4 m, width 2.5 m, height 4 m) up to its ceiling. The first layer (Stratum 3; c. 2 m thick) consisted mainly of stone debris that had fallen from the ceiling, mixed with eroded soils that had filtered into the cave over time. Stratum 3 sealed the remains of several fieldstone courses that lined and followed the cave’s contour, forming a semicircle that contained a layer of gray, powdery fill (c. 1 m thick), apparently the remains of a limekiln (Stratum 2). Although the limekiln is rather substantial, it was devoid of potsherds that could indicate its date. It may have been associated with the Hashmonean farmhouse, situated above and less than 10 m southeast of the cave. Stratum 1, badly affected by the limekiln activities, was composed of a fill (c. 0.5 m thick) deposited on the cave's bedrock floor, which contained anthropogenic materials, but no distinguishable features, such as pits or hearths. The ceramic assemblage from Stratum 1 was a mixture of Late Chalcolithic potsherds and several restorable, early EB I jars. It is yet too early to determine whether this mixture is a result of actual post-depositional processes in the cave or this assemblage is perceived as a mixture based on prevalent preconceptions concerning Late Chalcolithic and early EB I pottery assemblages. Among the finds was an intact, loop-handled jar, with clear affinities to early EB I rather than Late Chalcolithic, found, in situ, on the floor of the cave, as well as the indisputable fragments of a Chalcolithic churn, found higher up the fill. In the absence of any features, it seems likely that the cave was used for storage, rather than for dwelling in its earliest stages and perhaps for burial, based on a few, as yet unidentified bones.
Cave 3. Half of this cave was removed, including its original entrance. However, it was the most rewarding cave in Area C1. The cave (length c. 7 m, width 2 m, height 1.8 m) was filled with soil up to its ceiling. The top part of the fill consisted of natural, eroded soils that had filtered inside over the years. The lower fill (last 0.3–0.4 m), particularly in the southern part of the cave, rested directly on bedrock floor and yielded large amounts of well preserved, Late Chalcolithic pottery, a ring base fragment of a fenestrated basalt bowl, as well as over six hundred carbonized olive stones, retrieved through dry sieving of the soil; three olive stones were sampled for Carbon 14 dating. The concentration of olive stones, apparently stored away in the cave, lends additional support to the concept of cultivating olive trees in this area, as early as the Late Chalcolithic period. It joins another cache in the excess of 1200 carbonized olive stones that had previously been discovered in the nearby transitional Chalcolithic–early EB I site, excavated in the natural, deep depression between Hills A and B (HA-ESI 119). A functional relationship may perhaps be assumed between the abundant presence of olives pits and the hundreds of oval cupmarks, appearing on bedrock surfaces of Hills A–C.
Above the lower levels of the fill that contained the artifacts and olive stones, a stone wall, sealing off the south part of the cave and separating it from the north part, was built. Behind this wall, a deep man-made rectangular niche (0.6 × 0.7 m) was found, containing pottery that dated exclusively to early EB I. At a later stage, a deep cylindrical pit had been cut through the cave’s top in the north part, penetrating the bedrock floor and afflicting damage to the niche. This pit can be dated to the Persian period on the basis of pottery recovered from its fill, including several oil-lamp fragments.
Seven 4 × 4 m squares and two smaller probes were opened immediately south of the previously excavated Hashmonean farmhouse that comprised a central courtyard with a large oil press, surrounded on three sides by various rooms. The current excavation revealed two additional rooms to the west and a courtyard in the east, belonging to the south wing of the former complex (Fig. 2, center). Both rooms show three successive floor levels, as well as evidence of minor adjustments to several of their stone walls. The earliest floor level is dated to the late Hellenistic Period, namely the Hashmonean era (first century BCE) and the latest floor level––to the Roman period (beginning of the second century CE). A partially intact stone pavement in the north room had apparently been partly removed in antiquity. Inside this room and north of the remaining stone pavement an ash layer was noted, yielding a complete, soot-stained Roman-period juglet. This layer seems to indicate that the room was destroyed by fire and subsequently abandoned. A cylindrical shaft in the newly discovered courtyard led into a rectangular, carefully plastered water cistern, the second one found in association with the farmhouse. A small water channel, composed of two parallel alignments of small, standing fieldstones, led to this cistern from the south.
The pottery found in the rooms and courtyard indicates that the complex had been founded during the Hashmonean period and abandoned at the beginning of the second century CE, perhaps during the Bar-Kokhba Revolt.
It is noteworthy that wherever the earliest floor level was cut through, oval cupmarks, segments of stone walls and potsherds, dating to the Late Chalcolithic period, were found hewn into or built above bedrock. A wall segment of the farmhouse was built directly atop an earlier wall of small fieldstones that was, most likely, Chalcolithic in date and rested immediately above bedrock, which showed some shallow, oval grooves, undoubtedly of Chalcolithic date. In conjunction with the presence of the nearby Chalcolithic caves in Area C1, these in situ remains on bedrock point to intensive use of the area, as early as the Late Chalcolithic period.
A similar situation was noted when a square and deep rock-cut vat that adjoined the farmhouse and was perhaps originally associated with it, was excavated. The bottom of the vat was found cut through, leading into another karstic cave. The danger of possible roof collapse required the mechanical removal of bedrock above the cave, exposing its inner space (c. 2 × 6 × 7 m; Fig. 2, lower left corner), which was excavated down to bedrock. The most characteristic feature of the cave consisted of ten niches carved in the surrounding walls, some of which yielded large quantities of Late Chalcolithic pottery fragments. Several basalt vessel fragments and a single bone tool were also retrieved from the cave. This single thin layer of anthropogenic deposits (c. 0.1–0.2 m thick) is attested throughout the cave and indicates that it was used for a relatively short period of time, either for storage and/or as a dwelling. This layer rested on a c. 0.5 m thick fill of eroded and archeologically sterile soil mixed with limestone chips, which directly overlaid the cave's bedrock floor.
Near the southeast corner of the former farmhouse, a winepress that included a rectangular treading floor and a square collecting vat (Fig. 2, lower right corner) was revealed. The bottom of the collecting vat was also found cut through, leading into a deep shaft/pit, whose bedrock bottom was 4.5 m below surface. The shaft could be dated exclusively to early EB I on the basis of restorable potsherds recovered from the shaft's fill, including a storage jar with two pushed-up ledge handles.
A miqwe (ritual bath) excavated near the winepress was the second one identified in the area; apparently, it was associated with the farmhouse. The miqwe was accessed via a rectangular, almost shaft-like court (Fig. 2, extreme lower right corner) with a flight of shallow hewn stairs, descending toward the rectangular, slightly damaged rock-cut entrance to the miqwe. A number of earthen, plastered steps behind the entrance led down into an immersion pool, which originally was a karstic cave, adapted for its designated use as a miqwe. Various cracks in the cave’s walls were reinforced with fitting fieldstones in an attempt to prevent the possible collapse of the ceiling. The floor, roof and walls were coated with a blackish gritty layer, overlaid with a fine, water-proof, yellowish plaster. On the southern wall, a number of white-colored schematic representations of churches, alongside several apparent roads and road crossings, are depicted on a strip of the plaster (length 1.3 m, width 0.4 m; Fig. 3). To preserve for future research what appears to be a road map with holy sites, the painted plaster was eventually lifted from the wall by an IAA Conservation team. Perhaps the map had been drawn by a monk or pilgrim, staying in the abandoned miqwe during the Byzantine period, sometime in the sixth–seventh centuries CE, according to the pottery retrieved from the miqwe’s fill. A connection with the nearby Byzantine monastery/church at Horbat Hadat, less than 150 m south of the miqwe, seems plausible.
A rectangular, stone-cut passage was discerned a short distance south of the farmhouse. It led to a doorway of a burial cave, found blocked with a large boulder (Fig. 4). Several kokhim were visible in the cave, which could not be excavated due to IAA regulations. The cave dated, most likely, to the Hellenistic or Early Roman periods.
Cleaning bedrock surfaces and pockets in Area C2 yielded additional cupmarks or oval shallow grooves (Fig. 5), tentatively identified as stationary grinding facilities from the Chalcolithic period (HA-ESI 119), as well as bedrock pockets filled with soil that contained Chalcolithic potsherds and flints. A seemingly disproportionate high number of cornet bases and large lug handles with a triangular cross-section, which were friable and in a bad state of preservation, came from one such pocket.
Fifteen squares were opened over an area (c. 12 × 70 m) on the lower reaches of the hillside, to examine the various wall remains exposed by mechanical means and to uncover and clean bedrock surfaces in search of possible archeological deposits. The main feature in the area was an ancient, slightly curving road (length at least 60 m; S. Gibson, pers. comm.), oriented east–west and located on the very edge of one of the hillside’s terraces (Fig. 6). Integrated into the north side of the road was a long, straight segment of a possible retaining wall (length c. 6 m), built of medium-sized and dry-laid fieldstones, whose date is uncertain. The wall had only one defined face, on the southern side, facing uphill. Behind the wall and retained by it was a sloping fill of small fieldstones (c. 2 m wide) that directly overlaid bedrock and was the foundation of the road. The other side of the road, closest to the terrace edge, was not preserved. The majority of potsherds associated with this road dated from the Roman and Byzantine periods, providing an approximate date for its construction. Cutting through the road's stone foundation, several oval grooves/cupmarks of Chalcolithic date were exposed in the underlying bedrock, to the west of the wall segment. An unusually large Chalcolithic flint axe was found close to two oval cupmarks.
The only architectural remains in this area comprised the stone foundation of a circular structure (diam. c. 3 m; Fig. 7). The wall, built of a double row of medium-sized fieldstones and preserved a single course high, rested on a sterile layer of colluvial deposit that overlaid bedrock. A sunken, earth-beaten floor, slightly below the bottom level of the foundation course, yielded a spread of badly preserved early EB I potsherds, as well as an intact copper awl.
Prior to the excavation, small amounts of Chalcolithic potsherds had been noticed in three mechanically-dug trenches. Four adjoining probes (4 × 4 m) on the plateau, along the east slope of Hill C and overlooking Road 12, were excavated down to bedrock. A fifth 4 × 4 m probe c. 30 m to the southeast was excavated on the southern extreme of the plateau. The archaeological deposits appeared almost immediately below surface (c. 268 m above sea level) and reached a maximum depth of c. 1.5 m down to bedrock, depending on their location on the plateau that gently sloped in a northeast–southwest direction. A group of four, partly subterranean, silos in one of the probes was carved out of bedrock; one of the silos had been partially stone-lined. Three wall segments, founded directly on bedrock or on a sterile colluvial deposit, protruded from the balks of two probes and possibly belonged to domestic structures.
The remains exposed in the five probes are dated to the Late Chalcolithic period. The presence of cornets on the one hand and the absence of combed pottery, typical of the latter phases of Late Chalcolithic on the other, implies that the settlement remains on this plateau slightly predated the Chalcolithic remains excavated in 2004 (HA-ESI 119), which yielded combed pottery and only a few cornets.
The exposure of the southern wing of the Late Hellenistic-Roman (Hashmonean) farmhouse, which included an additional courtyard with two adjoining rooms, one of which had a stone pavement, a winepress and a collecting vat, another miqwe and a deeper, well-plastered water cistern, completed the plan of this complex.
A span of almost a century is envisaged for the continuous use of this complex, based on the identified floor levels in the two rooms.
The painted ‘road map’ on the south wall of the miqwe indicates a secondary use of the miqwe as a dwelling during the Byzantine period and provides an unexpected glimpse into the religious landscape of the period.
Around and below the farmhouse, bedrock is honeycombed with karstic caves and cavities, which were partially used for dwelling or storage during the Late Chalcolithic period; in one case the use extended into early EB I, whereas one other cavity was restricted to early EB I.
Subterranean storage facilities and wall segments of possible dwellings, all dating from the Late Chalcolithic period, were uncovered on the plateau.
The density of rock-cut grinding and pounding facilities evinces intensified usage of the area, presumably for agricultural activities, during the Late Chalcolithic period. This presumption, however, challenges the apparent marginal agricultural potential of this hilly, rocky area in terms of water resources and availability of quality arable land. Another possibly related point that still awaits clarification is the almost complete absence of animal bones in Late Chalcolithic contexts at the site.