Numerous excavations and surveys had been carried out at Tel Halif and its environs in the past. The most comprehensive is the Lahav Project, conducted by an American expedition, who exposed remains from the Early and Late Bronze Ages and from the Iron Age until the Crusader period, as well as remains dating to the nineteenth and twentieth centuries CE (The New Encyclopedia of Archaeological Excavations in the Holy Land 2, pp. 553–560; HA-ESI 120). Biran and Gophna excavated Iron Age burial caves south of the tell, not far from the current excavation area (Eretz Israel 9:27–39 [Hebrew]); Borowski excavated another Iron Age necropolis (Eretz Israel 23:13*–20*); A. Kloner exposed tombs from the Roman period and the beginning of the Byzantine period in Horbat Tilla, east of Tel Halif (Eretz Israel 17:325–332 [Hebrew]); and the American expedition excavated a cave that was adapted for dwelling in the late nineteenth–beginning of the twentieth century CE (Joe D. Seger. 1988. Location of the City of Ziklag in Light of the Tel Halif Excvations. In D. Urman and E. Stern, eds. Man and the Environment in the Southern Shephelah: Studies in Regional Geography and History. Giv‘atayim. Pp. 149–150 [Hebrew]).
 
Prior to the excavation, a cluster of six large adjacent caves (3–8; Fig. 2; Caves 7, 8 are not marked on plan) was identified, five of which (3–6, 8) were damaged. Squares were opened in Caves 3–6, which had been severely damaged and their ceilings were removed before the excavation commenced; only a partial excavation was undertaken, due to safety considerations. Cave 8 was examined after the excavation was completed, during antiquities inspection. It was difficult to determine if the caves were enlarged by quarrying due to the extreme damage caused to them. It seems that the caves, or at least some of them, were connected by narrow passages, but these could not be examined because of safety constraints. Two cisterns (1, 2; Fig. 2) were partially exposed south of the caves. The caves and cisterns are described below from north to south.
 
Cave 3
This cave (diam. c. 14 m; Fig. 3) was located northeast of the caves’ cluster. A trial square (4 × 7 m) was excavated in front of the opening, which was breeched by mechanical equipment in the southern part of the cave. The ancient remains were partly damaged by a modern refuse pit. A wall (W7; width 1 m), whose length is unknown and its general alignment is east–west, was exposed; it is possible that the wall extended across the entire width of the cave (up to 4 m long). The wall was built of small and medium limestone (0.15–0.30 × 0.20–0.50 m). The bedrock north of the wall was leveled and used as a floor. Fragments of pottery vessels dating to the Late Hellenistic period, including a black-slipped bowl with horizontal 'Pinched-Bow' handles (Fig. 4:1); a black-slipped base of a bowl (Fig. 4:2) and a cooking pot with an everted neck and handles (Fig. 4:3), were found in the fill above the floor (L147).
During the earthmoving work, a large ash level (L152; min. dimensions 4 × 5 m), which had been disturbed by mechanical equipment, was discovered on the floor of the cave. The pottery recovered from this level included an intact bowl with an inverted rim, characteristic of the Hellenistic period and fragments of red-slipped bowls from the Hellenistic or the Early Roman periods (not drawn). The rest of the finds were mixed and consisted of potsherds from the Hellenistic, Roman and Byzantine periods and the Middle Ages. It seems that the cave was used for habitation in these periods, particularly during the Hellenistic and Roman periods.
 
Caves 4 and 5 (Figs. 5, 6), located next to each other, were exposed 4 m west of Cave 3.
Cave 4. Two squares were opened in, the northern of the two: one square (4 × 5 m) in the southern part and the other (c. 2.8 × 4.0 m) at the northern end. Due to the extensive damage in the cave the remains were fragmentary; however, two or three phases of activity could be discerned in several places. A small cupmark (L160; upper diam. 0.25 m, depth 0.15 m; Fig. 7) was hewn in the qirton floor at the southwestern part of the cave. A level of ash mixed with brownish orange mud-brick material, tabun fragments and several potsherds (L133), which sealed the cupmark, was overlain with a crushed limestone floor (L127).
An installation built of two curved walls (W4, W6) was exposed on the bedrock floor, northeast of the cupmark. The large boulders, abutted by the walls from the south, could be collapse from the cave’s ceiling.Wall 4 (length 2.7 m, width 0.40–0.65 m, height 0.45 m) was built of a row of medium-sized limestone (0.10 × 0.34 × 0.40 m), while Wall 6 (length 1.9 m, width 0.5 m, height 0.3 m) consisted of small and large limestone (0.1 × 0.2 × 0.2–0.2 × 0.3 × 0.5 m).Between the two walls was a level of brownish orange clay that may have been a floor (L161). A section of a crushed chalk floor that abutted a bedrock surface (L162) was exposed east of W6.Both floors were overlain withlayers of fill, composed of gray soil mixed with small stones (L125, L128). The fill layers contained scant finds that consisted of potsherds and a few flint artifacts. East of Floor 125, close to where the mechanical equipment breeched an opening in the cave, was a series of thin ash and crushed chalk levels (floors?), interspaced with thick layers of fill that consisted of soil mixed with small stones, tabun fragments and potsherds (L115, L116, L118).
A massive wall segment (W5; length 2.8 m, width 0.7 m, height 1.8 m; Fig. 8), preserved to its full height up to ceiling of the cave, was exposed in the northern square. The wall was built of two rows of limestone and large flint stones that were roughly hewn (0.2 × 0.3 × 0.4 m), and a core of small fieldstones bonded in orange brown mud. It is possible that the wall was built next to the opening of the cave, but this is difficult to determine due the cave’s partial preservation. The wall was abutted from the west by a thin chalk floor (L143; Fig. 9) that was identical to Floor 127, exposed in the southern square. Floor 143 was placed on fill of light brown soil mixed with different size fieldstones (L144).Wall 5 and Floor 143 were covered with an especially thick fill layer (L132, L137) of stones and soil that contained fragments of pottery vessels and several flint artifacts.
The scant ceramic finds from Cave 4 were only potsherds. Except for the vessels from Ash Level 133, which dated to Iron Age II and included a red-painted, hand burnished fragment (Fig. 10:8), a handmade jar (Fig. 10:9) and a holemouth jar (Fig. 10:10), as well as a few body fragments of pottery vessels from this period in Fill 144 (not drawn), all the rest of the potsherds were found in mixed loci. These included a bowl from Middle Bronze IIB (Fig. 10:1); bowls from Late Bronze I (Fig. 10:2–4); jars from Late Bronze II (Fig. 10:5, 6); and two vessels dating to the beginning of the Iron Age II: a hand-burnished and slipped carinated bowl with a handle drawn from the rim (Fig. 10:7) and a jar with potter’s marks on its shoulder (Fig. 10:11). Other finds included body fragments of vessels that apparently dated to the Hellenistic period (not drawn), five flint artifacts from the Middle Bronze Age (below) and a few bones that were not in situ. These finds, alongside the installations and tabun remains in the cave, suggest that the cave may have been a dwelling, for at least part of the long time it was in use.
Cave 5. A single trial square (4 × 7 m) was openedin. Although the remains in the cave were fragmented due to severe damage, clean loci that abounded in artifacts had survived. Meager remains of two walls (W1, W2; benches?) built of brown mud bricks (c. 0.15/0.25 × 0.30 × 0.30 m) were exposed near the northern side of the cave, where the ceiling survived along the  opening that was breeched by mechanical equipment. Wall 2 (preserved length 1 m, width 0.4 m) was aligned north–south; Wall 1 (preserved length c. 1.2 m, width 0.4 m), generally aligned east–west, abutted W2 from the north; it probably delineated two spaces in the cave. Wall 1 consisted of a base of small stones superposed by mud bricks. It is apparent that the base was set on a level of ash that covered the floor of the cave, evidence of several activity phases in the cave. A segment of a wall (W3; length 0.6 m, width 0.5 m) built of small limestone fieldstones, was revealed near the northwestern end of W1; this wall may have separated Cave 5 from Cave 4. Collapsed stones and bricks were found along the southern wall of the cave.
Intermittent accumulations, probably remains of occupation levels that included mud bricks and ash concentrations, were exposed in the square: between W2 and the side of the cave (L113), east of W1 (L112, L117) and to its west, where three superposed levels (L119, L121/L124, L141) were uncovered; the ash level underlying W1 probably also belonged to this series of levels. In-situ concentrations of pottery vessels from the beginning of Iron Age II were found in Levels 112 and 113. Other finds from these levels included shells that originated in the Mediterranean Sea and small broken bones, including several human teeth.
Two levels containing mixed finds (L100, L114), probably due to surface proximity, were exposed close to the surface. The finds were ascribed to Middle Bronze II, Late Bronze II, beginning of the Iron Age and Iron Age II. Pockets of in-situ artifacts from the Iron Age were also found in Level 100, including an almost intact jar (see Fig. 15:1). Next to it was a bronze sword (length 0.7 m; Fig. 11) that has a flat blunt blade, a wide spine and a pointed tang with at least five nails in it; the date of this sword is unclear. Layers of fill that contained mixed finds were also discovered in a pit in the eastern part of the cave (L123), where the fill contained stones, inside the cave (L140, L149, L150), and along its southern side (L103). It is possible that the fill in Pit 123 and in the cave was placed intentionally as part of the preparatory construction in the beginning of Iron Age II, when the cave was cleared of most of its ancient finds.
 
The pottery from Middle Bronze II included open carinated bowls (Fig. 12:1, 2), a handmade cooking pot with a plastic decoration (Fig. 12:3), pithoi or jars with a molded rim and high neck (Fig. 12:4, 5) and a jug (Fig. 12:6). Three flint tools are attributed to this period (below). A deep bowl with a base ring (Fig. 12:7) and a cooking pot with an everted neck (Fig. 12:8) are ascribed to Late Bronze II. A deep carinated krater (Fig. 12:9) is attributed to the transition period between the Late Bronze Age and Iron Age I. Most of the pottery vessels from the beginning of Iron Age II were bowls (Fig. 13). With the exception of two (Fig. 13:6, 12), all are red slipped and hand burnished. The bowls included bell-shaped ones in the Philistine tradition (Fig. 13:10–12), one of which has a horizontal handle (Fig. 13:10), and several large deep bowls (fig. 13:13–15). Also attributed to this period are chalices (Fig. 14:1–3), one of which is decorated with red paint on its foot (Fig. 14:3); cooking pots (Fig. 14:4–8), including one with a curved carination and loop handles (Fig. 14:4) and a pot with an inverted neck and a ridge, from which two handles are drawn (Fig. 14:5); jars (Fig. 15:1–5), including a jar that was exposed almost in its entirety (Fig. 15:1) and a jar with a loop handle (Fig. 15:5); jugs (Fig. 15:6–8), one of which is burnished, with a grooved and slightly inverted rim and a long neck (Fig. 15:6); and a juglet (Fig. 15:9).A handle with a potter’s mark (Fig. 15:10), the likes of which are known, for example, from both Late Bronze Age and Iron Age strata at Tel Mor and Ashdod (IAA Reports 32, 2007, pp. 183–188), was also found.
The rich ceramic finds recovered from the cave, together with the sword, beads and human teeth suggest that it was used for burial at least during Iron Age II. It is possible that the benches were also used for interments, as a substitute for quarrying the bedrock, similar to the Iron Age burial caves exposed at other areas in Tel Halif.
 
Cave 6
The cave (diam. c. 10 m, height 3.25 m; Fig. 16), 4 m southeast of Cave 5, was found filled with stone collapse and light brown soil (L104). An ash level (L151) that covered a flagstone floor was discovered in a probe (2 × 3 m), opened in the western part of the cave. Fill 104 and Ash Layer 151 included a few potsherds from the Hellenistic and Early Roman periods (not drawn) and one flint implement from the Middle Bronze Age (below). It seems that the cave was used for dwelling in later periods.
 
Caves 7 and 8
These two caves were not excavated.Cave 7 was 11 m from the southern opening that was breeched in Cave 3. The cave was not damaged. Another cave (8) was exposed north of Cave 3, while overseeing the infrastructure work that was conducted upon completion of the excavation. The fill in Cave 8 included brown soil with small stones that covered an ash level. The finds gathered from the cave included fragments of bowls, cooking pots, jars, jugs and lamps dating to the Early Roman period (not drawn), indicating the cave was used for dwelling in this period.
 
Cisterns
The openings of two cisterns (1, 2; see Fig. 2), 10 m apart, were found 70 m south of the caves. An area (c. 8 × 10 m; Figs. 17, 18) was excavated around Cistern 1. The cistern opening was round (diam. 0.85 m) and bedrock-hewn and different size stones were arranged around it to the south, north and west. The cistern was rectangular (c. 3 × 5 m) and its sides were coated with thick gray plaster. An area (c. 6 × 8 m; Figs. 19, 20) was excavated around Cistern 2. The opening (0.55 × 0.75 m), on the eastern part of the cistern, was built of four courses (1.2 m high) of medium and large fieldstones; the cistern itself was not inspected. A few potsherds were found around the cisterns, including a black-slipped rim of a fish plate and a red-slipped krater rim, both dating to the Hellenistic period (not drawn). These finds may date the use of the cisterns.
 
The caves were used over a long period, from Middle Bronze II until Iron Age II, as well as in the Hellenistic, Roman and Byzantine periods. It seems that in ancient periods the caves were mostly used for dwelling, as well as for burial (Cave 5). Although the remains are fragmented, a connection is feasible between the site and the occupation strata at Tel Halif (Joe D. Seger 1988:139–150 [Hebrew]). Comparisons to the pottery assemblages recovered from the caves were found in Stratum X at Tel Halif, which was defined as a transition stratum between Middle Bronze IIC and the beginning of Late Bronze I, and in Stratum VIII (Late Bronze Age) and Stratum VII (end of Iron Age I and beginning of Iron Age II; late eleventh century–tenth century BCE) which included 'degenerated' Philistine pottery. An additional connection with Stratum VIII at Tel Halif is alluded to by the installation in Cave 4, built of a pair of curved walls (W4, W6) and similar to the round pits that were discovered in that stratum. The ceramic finds from the beginning of Iron Age II probably attest to ties with settlements in the Judean Shephelah, such as Tel Lakhish (Strata V and IV), Tel Batash (Stratum IV), as well as with coastal settlements, such as Ashdod, Tel Mor and Tel Qasile (Strata X and XI). These ties are also indicated by the finds in Stratum VII on the tell, which included ‘degenerated’ Philistine pottery.
The caves were used again as dwellings in the Hellenistic period, at the time when the cisterns were probably hewn. The caves continued to function as dwellings into the Roman period and to a lesser extent, later on, at least in parts that survived of them.
The use of caves as dwellings is known at different sites in the Northern Negev, such as Horbat Zak from the Roman and Byzantine periods and the Middle Ages (Y. Goren and P. Fabian. 1988. Horbat Zak—a Rural Settlement and Hiding Refuges from the Roman and Byzantine Periods. In D. Urman and E. Stern, eds. Man and the Environment in the Southern Shephelah, pp. 163–174 [Hebrew]) and Horbat Tilla, where evidence comes from the twentieth century CE (Joe D. Seger. 1988. Khirbat Khuweilifa: A Bedouin-Falachin Settlement on an Ancient Site. In D. Urman and E. Stern, eds. Man and the Environment in the Southern Shephelah, pp. 213–214 [Hebrew]).
 
The Flint Assemblage
Hamoudi Khalaily
 
The flint repertoire from Caves 3–5, 6 and 8 includes 391 artifacts, nine of which are tools and five are pounders made on flint pebbles. The frequency of the debitage (Table 1) and the absence of one of the debris components, such as chips, indicate that the retrieval of flint artifacts in the excavation was selective. Most of the flaked artifacts are fresh, without a patina, but most of them are also covered with a thick chalk cortex, indicating the flint originated in the soft limestone beds. Most of the raw material for the flint knapping in the caves derives from the Mishash Formation, which is frequent in the vicinity of the site. The quality of the flint is medium, its color is light gray and chalk veins are evident in it. Nonetheless, a few of the formal tools—the sickle blades, which were discovered in Caves 4 and 5—were made on high quality gray flint, which is not local. Since eight of the nine tools were found only in two of the caves (4, 5), the flint repertoire from these caves was examined.
It seems that the knapping was performed by non-professionals using available raw material inside the caves. This is evident from the way of shaping the cores, which were generally knapped using a hard hammer, aiming to produce only rough flakes. No systematic knapping technique was observed and most of the flakes were probably produced when the caves were being enlarged. Therefore, in light of the frequency of the flint components, the industry can be characterized as a flake industry: the proportion of flakes in the industrial debitage is over 60%, and most of the tools were also made on flakes. The blades are few and irregular. The only tool that was made on a standard blade is a sickle artifact from Cave 5. Another feature that characterizes the flakes is their size: more than two thirds of the flakes are fairly large; their average dimensions exceed 5 cm. Six of the seven cores are flake cores that were partially depleted. They are usually small and some of them have a thick chalk cortex. The debris contained mainly chunks, mostly natural.
 
Table 1: Frequency of the Industry’s Components
 
Primary Flakes
Flakes
Blades
Cores
Chips
Chunks
Tools
Cistern 1
2
13
1
2
-
6
-
Cave 3
2
3
1
1
-
1
-
Cave 4
12
46
7
1
1
50
5
Cave 5
33
68
4
3
28
51
3
Cave 6
8
22
-
-
3
13
1
Total
57
152
13
7
32
121
9
 
The Tools. Nine tools were discovered in three of the caves (Fig. 21): five in Cave 4, three in Cave 5 and one in Cave 6. No artifacts that bear signs of retouching were found in the rest of the caves or near the cisterns. The small number of tools in the assemblage makes it difficult to draw precise chronological conclusions regarding the habitation in the caves, because it seems that the fossile directeur—the sickle blades—penetrated from the surface and should not be related to the period when the caves were in use, although the tools types are common to the Middle Bronze Age.
Three tools were fashioned on blades of fine quality light gray flint; two of these are made on broad blades.One is a sickle blade made on a long blade (length 9.5 cm, width 2.7 cm, thickness c. 2 cm; Fig. 21:1). Its ends are truncated straight; it has a semi-abrupt back and its cutting edge is not retouched. The second artifact is a distal fragment of a geometric sickle blade (Fig. 21:2; Rosen, S.A. 1997. Lithics After the Stone Age. Walnut Creek, Ca). A truncation formed by abrupt retouching is evident at the end of the artifact, and its sides are not shaped; however, signs of use are apparent on the working edge, including pronounced sickle gloss on both the dorsal and ventral sides that probably stems from overuse. The third artifact is a short, broad sickle blade (Fig. 21:3). Its back is semi-abrupt and its ends are truncated. Retouching and prominent sickle gloss is visible on its cutting edge. One can reasonably assume that these artifacts were not prepared at the site; rather, they were brought to the cave as finished tools.
The rest are flake tools made of indigenous Mishash flint and all are ad hoc: three retouched flakes, one awl and a denticulated item.
Along with the flint tools, five flint pebbles were found; four are broken in the middle and one is complete (Fig. 21:4, 5). The pebbles are rounded and bear signs of having been worked: the cortex was removed by chipping and pattering. They were probably used as pounders, perhaps at the time when the caves were enlarged.
 
The group of tools includes two distinct indicators: (1) the presence of blades and broad sickle artifacts, characteristic of the post Early Bronze periods; (2) a geometric sickle artifact, the likes of which are known in both the Middle and Late Bronze Ages; however, it seems that this type is more common to the Middle Bronze Age. It therefore seems that the flint repertoire in Caves 4 and 5, where eight of the nine tools were found, can be dated to the Middle Bronze Age.