• from Ben-Ami et al. (2024:1)

This paper presents, for the first time, an analysis of the early Iron Age IIA occupation at ‘En Hazeva. A series of radiocarbon measurements from short-lived samples obtained from the site’s earlier occupation levels (Strata VIII–VII) were all dated to the 10th century BCE. It is noteworthy, that Stratum VII occupies the second half of the 10th century BCE exclusively, with its final phase around 900 BCE. Fixing the site’s absolute chronology has far-reaching implications, enabling the placement of the early Iron Age IIA settlement within the broader historical context. Situated c. 20 kilometres from the most significant copper industry centre in the Levant — Khirbet en-Nahas, ‘En Hazeva enjoyed a strategic location in the transport network of copper through the Negev Highlands and the Beer-Sheba Valley to the Mediterranean seaports. It is claimed that the economic prosperity related to copper production at Khirbet en-Nahas during the early Iron Age IIA was ‘En Hazeva’s raison d’être.

• from Ben-Ami et al. (2024:2)

... The Iron Age is represented at ‘En Hazeva by three strata — VIII, VII and VI, ranging from Iron Age I to the Iron Age IIC (the Strata designation follows the new stratigraphical scheme that resulted from the preparation of the material for final publication). The earliest remains, dated to the Iron Age I (Stratum VIII), were retrieved mainly during the 2013 and 2023 excavations (above), which took place long after the Cohen-Yisrael project ended. They include a few architectural remains and related finds recovered in a limited area, stratigraphically located beneath the remains of Stratum VII and differing from them significantly in magnitude (this paper focuses on the early Iron Age IIA at ‘En Hazeva (Stratum VII). Stratum VIII’s remains were barely excavated and await more significant exposure, which is planned for the near future). The early Iron Age IIA remains at ‘En Hazeva, the focus of this paper, are primarily represented by a single structure — Building 3011, which displays a well-planned, skilfully executed architecture. The comprehensive study of the available material, including the 2023 excavation results, enabled radiocarbon dating and setting the building in the broader cultural sphere.

• from Ben-Ami et al. (2024:2-11)

Remains belonging to Stratum VII were exposed below the Stratum VI fortress’ gate complex, which dates to Iron Age IIB–C. Given the spring nearby, the concentration of remains in this part of the site should not be considered coincidental. The remains ascribed to Stratum VII include Building 3011, as well as some wall segments and installations exposed nearby. Two stratigraphic phases were discerned in the building (Strata VIIB and VIIA); the later phase consisted mainly of raising the floor levels in most rooms, and some architectural alterations (Figs 2–3)⁵.

Building 3011 features a massive substructure. Despite being partly damaged by the construction of the immense fortress over it in the subsequent stratum, its foundation walls are relatively well preserved and stand to a considerable height, some measuring over 2.5 m. No foundation trenches for the substructure walls were distinguished, and this observation further emphasises Building 3011’s construction technique. The substructure followed the ‘built-up foundation walls’ (Ussishkin 1980: 10–11; 2004: 510–12). Accordingly, the foundation walls’ lower courses were built directly on the surface of the time, scarcely penetrating the earlier level. Short-lived organic samples retrieved from beneath the building’s southern wall (W871) were submitted to radiocarbon dating (RTD 12153, see below). As the foundation walls were gradually raised, fills were laid horizontally, at intervals, to fill in the spaces between the walls. This procedure was repeated until the internal space delineated between the foundation walls was filled to the desired height. Only then were the floors in the various rooms laid, directly over the extensively accumulated constructional fills. Evidently, the high foundation walls and massive constructional fills were intended to raise the structure’s living floors above its surroundings.

The constructional fill, consisting of two distinct units, was almost devoid of artefacts and stones (Fig. 4). The lower unit is a dark fill, while the upper is a thick layer of whitish-yellowish fill. The two fill units are easy to distinguish, indicating deliberate planning and execution. Constructional fills were also traced along the outer face of the building’s southern and eastern walls (W871 and W340, respectively); presumably, they surrounded the substructure walls on all four sides. These fills formed diagonal layers slanting at 45° from the outer face of the building’s foundation walls, toward the extant surface (Figs 5–6). On the eastern side, these earthen fills (L3238 + L3008+L3228+L127) were clad with stones on their lower outer end (W887, W706, W707). A thin blackish layer of ash, sandwiched between two hard, compact layers of white-coloured plaster, coated the southern side. The purpose of these coatings was to stabilize the sloping fills. The constructional fill created a podium-shaped substructure. As this construction process found similar reflections in the six-chambered gate at Tel Megiddo, it is appropriate to quote Ussishkin’s methodological observation (1980: 10):

Similar fills [constructional fills — DBA], some times ramp-shaped, were usually laid against the outer face of the foundation walls, to support them from outside. This way, the foundation walls are initially constructed as free-standing walls, and once stripped of the supporting constructional fill they would look like typical walls of the superstructure of a building. Furthermore, once the built-up foundation walls of the whole structure are exposed to their supporting constructional fill, they look like a free-standing structure built on the surface of the previous stratum.

The foundation walls consisted of small and medium-sized stones, with some larger ones integrated into the building’s corners. The stones are unworked, sometimes only roughly hewn. Many are dark greyish flint stones of the Mishash Formation, others are brighter in colour, and belong to the Hazeva Group and the Taqiye Formation.⁶ No ashlars were used in any of the walls. These characteristics make the Stratum-VII walls distinct from those of the overlying Iron Age IIB–C fortress in the subsequent stratum. The latter was constructed solely of whitish-yellowish limestones. This choice of building materials for the substructure and superstructure of Building 3011 makes the distinction between the Stratum VII architectural remains and those of Stratum VI easy to follow. In some instances, the sun-dried mudbricks that formed Building 3011’s superstructure were discovered in situ; placed on top of the substructure’s stone walls. These mudbricks were of orangish-brownish and greyish hues and had average dimensions of 0.3 ×0.5×0.1 m. Short-lived organic samples retrieved from these mudbricks were sampled for radiocarbon dating (see below).

The building is rectangular, nearly square in plan (c. 13.0 × 12.0 m). It consists of five rooms: Room 5 in the centre, surrounded by Rooms 1–4 (Fig. 7). The entrance to the building was not exposed, but it is likely to have been located on the eastern wall (W340), facing the spring. Three doorways that connected Room 5 with three surrounding rooms in the north, east and south were excavated (Rooms 4, 3 and 2, respectively). The south-western side of this room and the southern part of Room 1 were consider ably damaged by the white mudbrick installation (L3137) that was constructed atop Building 3011 in the subsequent stratum (VI). Since this is the first time the building is being published, it deserves a brief description.

• Room 1 is an elongated space (9.5 × 2.5 m), located on the western side of the building. As noted, the southern side of the room was damaged by the later mudbrick installation built over it. To make room for the installation, the builders of the subsequent stratum had to dismantle some walls on the south-western side of Building 3011, mudbrick superstructure and stone substructure alike. Hence, W871 and W872 were poorly preserved here, and the southern part of W859 was almost entirely razed (see below). A partition wall might have divided this elongated room. No floor level was detected in the limited area not occupied by the Stratum VI installation.


• Room 2 is the smallest in the building (4.5 × 2.5 m), located on its southern side. The room’s western boundary was cut and built over by the Stratum VI mudbrick installation. An entrance set in the north-eastern corner of the room connected it with the central Room 5. The floor features a stone-paved foundation (L3217), slightly slanting eastward toward the doorway.


• Room 3 (6.0 ×2.5 m) occupies the eastern side of Building 3011. Mudbricks were preserved in situ along W855 on the west, and W841 on the north. The single entryway into this room was set in W332 +W855, connecting it with Room 5. The entrance is flanked by two mudbrick doorposts, preserved to a height of 1 m above the floor. During the early phase (VIIB), a compact earth floor (L130, excavated in 2023), abutted the threshold. Hand-made pottery (mostly referred to as ‘Negebite ware’), including a complete hand-made juglet, was found on this floor along with wheel-made pottery. A thick, circular layer of dark ash (L129, excavated in 2023) was found in the north-western corner of the room. This ash material is evidence of a circular installation whose walls were almost entirely dismantled in the subsequent phase. The ash contained charred date pits that yielded a radiocarbon date in the later part of the 10th century BCE (RTD 12154, see below). Fragments of a hand-made vessel lay in the ash.
  
  During Phase VIIA, the floor level was raised by c. 0.5 m, and a new compact earth floor was added above the previous one. A single flagstone abutting the northern doorpost was preserved in situ in the later threshold. A circular installation (L1699) was set in the northern part of the room. The installation was constructed of whitish-greenish clay. Similar installations were also excavated in Rooms 4 and 5 (below). Ash layers were excavated inside and next to the installation. A complete hand-made cooking krater was found on the floor close to Installation L1699. A narrow mudbrick wall (W873) was built perpendicular to the doorpost in W855. This thin mud brick wall stretched eastward, serving as a short partition wall in Room 3. A compact earth floor, L3126, was excavated in the southern part of this room. A semi-circular installation, L3130, similar to L1699 above, was unearthed just south of the partition wall W873, leaning against the eastern face of W855.


• Room 4 is a large rectangular space (8.0 × 2.6 m) that occupies the northern side of Building 3011. A stone-built threshold connecting Room 4 with the central Room 5 was unearthed at the eastern end of W330. The original floor (Phase VIIB, L1620b), visible in the section that cuts into the room’s eastern side, rests directly over the massive constructional fill mixed with mudbricks and mudbrick material. In the later phase (Phase VIIA), the floor level was raised by c. 0.65 m (L1620a + L1610), as it was in the adjacent Room 3. Furthermore, additional features noted in the later phase of Room 3 also characterize Room 4’s latest phase. These include a circular installation (L1629) and a thin curvilinear mudbrick wall (W331). Among the finds worth noting ascribed to this floor are ostrich eggshell fragments.


• Room 5 (4.7×3.2 m) is situated at the centre of the building, surrounded by the other four rooms, and with an entryway to each. It is likely that this room served as an open courtyard. Its original floor (Phase VIIB, L1658b) abutted the stone thresholds of the surrounding rooms. This compact earth floor rested, like the building’s original floors, on the massive constructional fills whose upper part consisted of a fill layer mixed with mudbrick material. In the subsequent Phase VIIA, the floor level was raised (L3011). A circular installation, similar in character to those found in the adjacent rooms (above), was set close to the doorway leading to Room 4. Dozens of charred date pits were found in the ash layer in and around the installation (Fig. 8). These were radiocarbon-dated to the last quarter of the 10th century BCE (RTD 12207, see below).

In addition to the constructional fills that slope from the outer face of the foundation walls to create a ramp-shaped platform (above), some wall stubs and an oven excavated outside Building 3011 could be safely ascribed to Stratum VII. These were built atop the constructional fill, outside the building near its eastern wall (W340). These walls were built, without exception, from fieldstones similar to the substructure of Building 3011.

Analysis of the Stratum VII remains brought to light two significant architectural and stratigraphical observations. Firstly, the massive fills that create an artificial, platform-like foundation for Building 3011’s substructure testify to a considerable investment in workforce and to close acquaintance with large-scale building enterprises. To the best of our knowledge, this phenomenon, of using elevated built-up substructures that integrate massive constructional fills, has been noticed primarily in monumental public structures. Prominent examples are the four- chambers city gate at Megiddo (Ussishkin 1980) and the city gate and palace at Lachish (Ussishkin 2004). Indeed, Building 3011 does not integrate ashlar masonry in its substructure as did the other examples, and there is a clear tendency to use local raw materials; nonetheless, the exceptional endeavours characterizing its construction distinguish it from the common domestic structures of its type and allude to an organized initiative (below). Moreover, the construction of the immense fortress (Stratum VI) in Iron Age IIB occurred while some of Building 3011’s walls still stood to a considerable height. The archaeological evidence testifies to this building having been covered by the fortress’ constructional earthen fills, and these fills are the main reason for its excellent state of preservation, including some of the mudbrick superstructure walls. The only damage occurred to the walls that underlay Stratum VI’s foundations. This is best reflected in the Stratum-VI mudbrick installation L3137, which was built over the south-western side of Building 3011 (above). It is only the understanding that L3137, with its thin mud brick walls, was an underground facility (and therefore well-protected), while Building 3011’s inner walls formed part of its superstructure, that put these elements in their proper context. The stratigraphical and architectural relationships between these elements have important implications for evaluating the diachronic circumstances at the final phase of Building 3011 (and Stratum VII overall) and the construction of the overlying Stratum VI fortress. Accordingly, the scarcity of finds in general, and of pottery in particular, on its floors suggests that the inhabitants left the site by choice, taking their portable belongings and valuables. The immense fortress was constructed in the Iron Age IIB, long after Building 3011 had been abandoned, and a considerable chronological gap separates the two.

One of the primary objectives of this project is to establish high-resolution radiocarbon dating for the ‘En Hazeva Iron Age sequence. The first renewed excavation season at ‘En Hazeva, in 2023, allowed sampling of several secure and stratified contexts that provided high-precision radiocarbon dating for Building 3011. Setting the absolute chronology for ‘En Hazeva Building 3011 has far-reaching implications beyond the site’s occupation history, for it sheds light on the early Iron Age IIA settlement in the broader geographical and historical context.

• from Ben-Ami et al. (2024:14-17)

The earliest radiocarbon measurement obtained from ‘En Hazeva dates to 980–950 BCE and is associated with Stratum VIII. The three short-lived samples analyzed from this early context, which is stratigraphically located beneath the Stratum VII floors, point to a date in the Iron Age I for Stratum VIII. Thus far, we have only measured its latest upper boundary; our project aims for a broader architectural and stratigraphical exposure of Stratum VIII in the coming excavation season, which will undoubtedly allow us to study its character and set the absolute date of its foundation phase.

Building 3011 is a nearly complete architectural unit. Structures with this plan are common farther west, in the Negev Highlands, during the early Iron Age IIA (second half of the 10th century–first half of the 9th century BCE; see Finkelstein and Piasetzky, 2010; Herzog and Singer-Avitz 2004). Cohen and Yisrael, who excavated the building, suggested that it was part of a fortress. According to them, the early Iron Age IIA ‘fortress’ at ‘En Hazeva was part of a fortress network established in the 10th century BCE under royal initiative. They ascribed it to the United Monarchy during Solomon’s reign. The reasons behind establishing these ‘fortresses’ were twofold — to safeguard the roads crossing the central Negev, and to defend the kingdom’s southern border. Cohen and Yisrael (1995: 232; 1996: 91; and see below) attributed the destruction of the ‘fortresses’, ‘En Hazeva included, to Sheshonq’s military campaign in c. 926 BCE. However, the data retrieved during the 2023 excavations at ‘En Hazeva lead to a somewhat different interpretation of the fortress’ precise date and character.

Recent investigations into the early Iron Age IIA settlements in the southern arid regions of southern Israel and southern Jordan have provided significant insights that allow ‘En Hazeva Stratum VII to be set in a broader historical context. The ‘En Hazeva building parallels in plan several Negev-Highlands structures, the closest parallel being Har Hemet (Cohen and Cohen-Amin 2004: fig. 95:6). Like the Negev-Highlands structures, the ‘En Hazeva building included a pottery assemblage dated to the early Iron Age IIA, comprising two wares typical of the region — the wheel-made and the hand-made (i.e., ‘Negebite’) vessels (Cohen and Cohen-Amin 2004: 121–41, 6*–8*).⁷ Thus, based on its chronological setting, architectural plan and pottery assemblages, Building 3011 is associated with this early Iron Age IIA settlement wave (see below).

Although it is widely accepted that the Iron Age settlements in the Negev Highlands should all be dated to the early Iron Age IIA (Finkelstein 2014: 95; Martin et al. 2013), other fundamental issues remain in dispute. Debates are mainly concerned with the ‘who’ and ‘why’ questions, and relate primarily to the identity of the inhabitants, the reasons for their sudden establishment in this arid zone, and their abrupt demise. As these subjects have been widely discussed in literature without reaching a consensual view, it is futile to repeat the various approaches and sets of arguments here (for detailed discussions and references, see Aharoni 1967; Cohen 1979; Cohen and Cohen-Amin 2004; Eitam 1988; Faust 2006; Finkelstein 1984; Haiman 1994; Herzog 1983; Martin et al. 2013; Meshel 1994; and Rothenberg 1967).

Many of the studies have shown that the incentive for the Negev Highlands settlements was economic, since they chronologically coincide with the peak of copper production activity at Faynan (Bienkowski 2022: 128; Howland 2021: 75–79). Intensive research into the Arabah copper production centres, primarily the leading site of Khirbet en-Nahas, concludes that copper production lasted from the 12th to the 9th centuries BCE, with major mining activity occurring during the early Iron Age IIA (Ben-Yosef 2010; 2018; Ben-Yosef and Thomas 2023; Levy 2009). In this respect, the significance of ‘En Hazeva’s geographical proximity to the western end of Wadi Faynan cannot be overestimated. Situated close to the Wadis Faynan-Arabah meeting point, c. 20 kilometres, only, from the largest copper industry centre in the Levant, Khirbet en-Nahas (Ben-Yosef et al. 2014: 544 and figs 6.38–6.39; Levy et al. 2004: 867), ‘En Hazeva enjoyed a prime location in the copper traffic network associated with the two copper production regions, Timna in the southern Arabah Valley and the close-by Wadi Faynan. Since much of the Arabah copper had to cross to the Negev Highlands and proceed through the Beer-Sheba Valley to Gaza and the Mediterranean seaports, ‘En Hazeva was a strategic station along the commercial trade routes (Finkelstein 2014: 96; Singer-Avitz 1999: 10; 2008: 79). Its strategic location in this international copper trade network is further supported on petrographic grounds: preliminary results of the petrographic analyses, carried out on the pottery assemblage from Building 3011, testify that the hand-made ware included crushed copper slag as a tempering agent in the clay matrix (see below). This phenomenon aligns perfectly with studies establishing a connection between the Negev Highlands and the Arabah copper production centres (Howland 2021: 70–72 with extensive literature). Thus, the Negev Highlands phenomenon must be seen as a westward expansion of the activities centred in the Arabah Valley (Ben-Yosef 2023: 240; for the resemblance of the Area R building at Khirbet an-Nahas to some of the Negev Highlands structures, particularly that of Atar Haroa, see Bienkowski 2022: 126–27; Levy et al. 2014a: 231–32), with ‘En Hazeva located close to its hub. The absolute chronology of Building 3011 is harmoniously aligned with the peak recorded in the Arabah mining activity during the early Iron Age IIA. This economic prosperity was ‘En Hazeva’s raison d’être during this period.

To assess the overlapping activity intensities at ‘En Hazeva with the two bounding regions — Faynan on its east and the Negev Highlands settlements on its west — the radiocarbon dates from Khirbet en-Nahas and the Negev Highlands site of Atar Haroa were compared. One hundred and ten measurements were published from Khirbet en-Nahas (Levy et al. 2010; 2014a; see Tebes 2022: n. 4), and 16 from Atar Haroa (Boaretto et al. 2010). In principle, a site-specific stratigraphic model is the ideal approach to compare site chronologies. However, Atar Haroa has no discernible stratigraphy and is considered to be a single-layer site. In the case of Khirbet en-Nahas, the measured samples, including charred wood and charred seeds, are from stratigraphic layers in different areas, meaning that several chronologies would need to be built, which is beyond the scope of this study. Consequently, the Sum function available at the OxCal software was used to obtain a rough estimate of the dominant activity phases at these two sites and compare them with ‘En Hazeva. The individual probability distributions of all the samples measured at each site were combined into a single probability distribution plot (Fig. 17). These plots illustrate the timespan in which the sites show high values of occupation intensity. It should be taken into account that the resolution of these plots is dependent on the number of samples that the excavators decided to date, particularly at Khirbet en-Nahas. Nevertheless, they give the trends of occupation intensity. It should be noted that since most of the samples from Khirbet en-Nahas were of charcoal, there is a potential bias towards slightly older dates; however, sampling outer rings reduces the ‘old wood effect’. On the other hand, at the site of Atar Haroa, the samples were of short-lived material.



These comparisons, between the three sites, allow for some conclusions to be drawn. It is clearly shown that ‘En Hazeva was already inhabited during Iron Age I (Stratum VIII), a time when substantial activity was noted at Khirbet en-Nahas. Building 3011, of the subsequent Stratum VII at ‘En Hazeva, is dated to the second half of the 10th century BCE, coinciding with increased activity at Khirbet en-Nahas. However, while the peak of the Faynan copper industry seems to be around the mid-9th century BCE, corresponding to the chronology of Atar Haroa, En Hazeva was abandoned slightly earlier, at the close of the 10th century BCE (see below).

The alterations recorded in Building 3011 distinguish ‘En Hazeva from most other Negev Highlands settlements, which, in general, lack complex stratigraphy, and mostly feature one short-span occupation phase (e.g., Eitam 1979: 128; Faust 2006: 139; Haiman 1994: 57–58; Meshel 1994: 58). Building 3011 provides a firsthand opportunity to pinpoint the dating of its establishment, duration and final use. The stratigraphically well-controlled radiocarbon dates of the sampled mudbricks, offer a range between 950 and 920 BCE, setting the mid-10th century BCE as the earliest possible date for its construction. The date pits sampled from the floors of the two phases put Phase VIIB at c. 940–910 BCE, while Phase VIIA is dated to c. 935–900 BCE. Hence, Building 3011 is dated exclusively within the early Iron Age IIA, occupying the second half of the 10th century BCE. These stratigraphical and chronological settings have significant implications for the historical circumstances in the southern arid regions, namely the Arabah Valley and the Negev Highlands, during the early Iron Age IIA.

Historically, the second half of the 10th century BCE is primarily known for Pharaoh Sheshonq I’s (biblical Shishak) Asiatic campaign to Canaan (for some uncertainty regarding the exact year of Sheshonq’s campaign, see Boaretto et al. 2010, n. 4; Fantalkin and Finkelstein 2006; Finkelstein 2002; c.f., Webster et al. 2023). The Sheshonq I list, on a wall of the temple of Amun at Karnak, boasting about his campaign into Canaan, includes a large group of toponyms widely accepted as being located in Canaan’s southern, arid zone (Kitchen 1986: 293–300, 432–47; 2003; Na’aman 1998 with further references). Archaeologically, this region is known to have experienced a sudden wave of short-term settlements during the early Iron Age IIA, and the two (i.e., the campaign and the settlement wave) demonstrate, to a certain extent, possible contemporaneity (Mazar 2007: 148). It is widely accepted that this settlement wave was economically motivated by the Arabah copper production and commerce system. Indeed, recent works focusing on the economic prosperity that evolved from the copper exploitation system of the early Iron Age IIA in the southern arid zones have already suggested a link between this settlement wave and Sheshonq’s enterprise in southern Canaan, stressing the Pharaoh’s intention to gain control over the copper production and trade system and improve its efficiency (Boaretto et al. 2010: 10; Fantalkin and Finkelstein 2006: 27– 28; Martin et al. 2013: 3790; c.f. Ben-Yosef 2010: 971–77; 2016: 193; 2023: 240, 247; for a scarab of this Pharaoh found at Wadi Fidan see Levy et al. 2014b). This interpretation is supported by a set of radiocarbon measurements from short-lived samples measured at Atar Haroa (Boaretto et al. 2010) and similar determinations from the nearby site of Nahal Boqer (Finkelstein 2014: 95, 98; Shahack-Gross et al. 2014: 107–08; 114–15) alluding to early Iron Age IIA dates in the 9th century BCE. In light of the new radiocarbon chronology of ‘En Hazeva, these ¹⁴C determinations invalidate the possibility that the settlements were destroyed by, or even abandoned as a result of, the Egyptian Pharaoh’s campaign (c.f., Cohen and Cohen-Amin 2004: 157; Cohen and Yisrael 1995: 232; 1996: 91; Faust 2006: 153; Mazar 1997: 160–61; 2007: 149–52; 2010: 31; Na’aman 1992: 83–85, 88; 1998).

‘En Hazeva provides a unique insight into this complex period due to its stratigraphical, architectural and chronological settings, and the fact that it was not destroyed but rather abandoned in the late 10th century BCE. If indeed the ultimate candidate for royal initiative in the Arabah Valley and the Negev Highlands during the early Iron Age IIA is Pharaoh Sheshonq I, then Building 3011 resonates with this scenario. As noted earlier, Building 3011 was founded on a massive elevated built-up substructure, a fact that attests to a significant investment in labour and to knowledge of complex construction techniques. Initiatives on such a scale must have been carried out by an authority with an established tradition of large-scale construction operations, which, therefore, makes the local desert inhabitants less likely candidates (for a different view, see the ‘architectural bias’ in Ben-Yosef 2010: 988; 2019; 2020; 2023; Ben-Yosef and Thomas 2023). Indeed, the coarse hand-made pottery found in the building, and understood by many scholars as a product of the desert nomads, suggests that the building has pastoral nomadic traits, possibly of the local desert population (Faust 2006: 149; Herzog 1984: 26; Martin and Finkelstein 2013: 12; Meshel 1994: 59; Tebes 2006). Nonetheless, this does not mean that this section of the population was, necessarily, responsible for the planning and execution of the enterprise. Any attempt to claim that groups with nomadic back ground were involved in the construction of the site, or to consider them among its dwellers, solely on the basis of the hand-made ware in the pottery assemblage, would, at best, be mere speculation. Hand-made ware has been accepted as a distinct regional phenomenon, confined almost exclusively to the desert regions of the Negev Highlands and Wadi Arabah. Recently, however, this view has changed, and it is now deemed to have been transported into these regions from Wadi Faynan via the movement of people (Bienkowski 2022: 127; Martin et al. 2013; Martin and Finkelstein 2013; Yahalom-Mack et.al 2015). Thus, the hand-made vessels cannot be considered a firm indicator of either the participation of the desert communities in the construction of Building 3011, or of their actual presence there (see Bernick-Greenberg 2007: 210); it can only allude to a certain degree of contact between the building’s dwellers and this nomadic element.

• from Ben-Ami et al. (2024:17-18)

The new high-resolution dates from stratified contexts enabled an absolute chronological setting for the early Iron Age settlement at ‘En Hazeva. The above analysis testifies to the special status that should be assigned to ‘En Hazeva during the early Iron Age IIA; a status that derived directly from its role in the Arabah copper trade network. Due to its prime location at the crossing of the Arabah (see also Ben-Yosef et al. 2014: figs 6.38–6.39), ‘En Hazeva became a strategic waystation along the commercial network transport ing the Arabah Valley copper to the Mediterranean seaports, with the residents of ‘En Hazeva working under the auspices of the authorities and holding administrative positions primarily relating to Arabah copper production and commerce (see already Mazar 2007: 151). Given that the aim of Sheshonq’s campaign was to control the Faynan copper exploitation and trade system, and to redirect the copper trade from the eastern highway to the coastal plain and Egypt (e.g., Finkelstein 2014: 96), Egypt becomes the ‘usual suspect’ for initiating the construction of Building 3011 at ‘En Hazeva.

The site’s abandonment is in keeping with the abandonment of many other sites in the region at about the same time and should, therefore, be viewed against this more comprehensive background. It was a gradual process, tightly associated with the gradual decrease in Arabah copper production towards the late 9th century BCE. While the end of this process is due to the revival of the import of Cypriot copper to the Levant in the late Iron Age IIA, the abandonment of Building 3011 at ‘En Hazeva could have resulted from Egypt losing its firm grasp over the region following Sheshonq I’s sudden death. During the time of Sheshonq I’s successor, Osorkon I, Egypt kept its foreign affairs policy unchanged, maintaining its alliance with the commercial port of Byblos. But it was not too long before internal instability struck, preventing the Egyptians from establishing a colonial system in the southern Levant during the Third Intermediate period. These troubles culminated during the reign of Osorkon II (whose cartouches were found on a large alabaster vessel at Samaria), when Upper Egypt and the Thebaid sought greater independence, challenging Osorkon II’s effective, absolute rule, and thus rendering any foreign adventure beyond Egypt’s capability (Kitchen 1986: 324). It seems, however, that the early years of the 22nd dynasty were nothing more than a ‘short-period revival of Egypt as a key player in the Levant’ (Finkelstein and Piasetzky 2008: 89). If the Egyptian interest in controlling the southern Levantine copper trade network was the raison d’être for ‘En Hazeva Building 3011, its loss of influence in the region eventually led to the building’s doom.

Building 3011 stood abandoned for nearly a couple of centuries showing no trace of later occupation activity. The immense fortress established at ‘En Hazeva over the abandoned building (Stratum VI), opens a new chapter in the site’s history, dating to the Iron Age IIB–C. The fortress’ massive constructional fills buried Building 3011 entirely, and were the main reason for the latter’s excellent preservation, including its mudbrick superstructure walls.

• from Cohen and Yisrael (1995:230-231)

... There are several good candidates for the builders of the Stratum 5 fortress. The results of the most recent excavations at the site have contributed to a change in our thinking concerning the initiator of this construction project. We now believe that it may have been built in the ninth-eighth centuries BCE rather than a century later as previously suggested. A look at the relations between Judea and Edom as they are described in the Bible is necessary to understand who the possible architects were and how we have arrived at our choice of the most likely candidate. cEn Haseva Stratum 5 may represent a military base that was enlarged as necessitated by the political climate of the times. The initial early phase, the gate complex, may have been constructed by Jehoshaphat (867-846 BCE) when "there was no king in Edom, a deputy was king" (1Kgs 22:48). 1 Kgs reports that, in an unsuccessful attempt to repeat Solomon's achievements, "Jehoshaphat made ships of Tarshish to go to Ophir for gold, but they did not go, for the ships were wrecked at Ezion-geber" (1 Kgs 22:49; Bartlett 1989:115-6). This attribution would also find support in 2 Chr 17:2: 'And he [Jehoshaphat] placed forces in all the fortified cities of Judah, and set garrisons in the land of Judah." Later, perhaps at the end of his reign, the fortress was enlarged to accommodate the Israelite/Judean retaliatory campaign against Mesha, king of Moab (mid-ninth century BCE; 2 Kgs 3:4-15), who mentions his rebellion against the king of Israel in his Stele (Bartlett 1989:116-22; Dearman 1989). The large Stratum 5 fortress may have served as the deployment center for this invasion. This would not only lend credence to the biblical statement that Jehoshaphat built fortresses and storage cities in Judah (2 Chr 17:12), but would also serve to strengthen the identification of the magazines at CEn Haseva as, in fact, a storeroom complex and not a building of some other kind (e.g., stables, barracks, or market places). The similarity in plan between the Stratum 5 fortress at CEn Haseva and the Iron Age fortifications at Tel Jezreel (Ussishkin and Woodhead 1994) suggests that they may date to approximately the same time, i.e., the ninth century BCE or some time thereafter.

Another possibility is that the fortress was built by Amaziah (798-769 BCE), the son of Joash, who was diligent in fortifying his kingdom both from within and without, and, after instituting reforms in the army, went to war with Edom. He defeated the Edomites in the Valley of Salt, in the northern Arabah, and then went on to conquer Sela (2 Kgs 14:7), renaming it Jokthe-el, and settling descendants from the Tribe of Judah there. Was it from this fortress that he set out against the Edomites? Or was the fortress built during the reign of Amaziah's son Uzziah (769-733 BCE; 2 Kgs 15:1), the powerful and active king who "built Elot and restored it to Judah" (2 Chr 26:2; 2 Kgs 14:22;), fortified the borders of his kingdom "and built towers in the wilderness" (2 Chr 26:10), and, like his father before him, strengthened his army in order to prevail over his enemies (2 Chr 26:13)? Although there is a strong possibility, based on archaeological, historical, and biblical considerations, that Jehoshaphat was the builder of the Stratum 5 fortress, there are, nevertheless, archaeological remains dating to Uzziah, and a case may be made to support his having engineered at least part of the construction of this fortress.

We should also take into consideration the possibility that the Stratum 5 fortress was constructed following an Assyrian takeover similar to that which occurred in Edom during the reign of Adad-nirari III (810-783 BCE) and for which there is evidence in an inscription from Calah (Pritchard 1969:281; Bartlett 1989:124; Ephal 1984:76; Millard 1992). The inscription contains the first reference to Edom as a kind of vassal state, paying tribute to the king of Assyria; subsequent references abound (Bartlett 1989; Ephal 1984). Assyria's intense interest in protecting, if not controlling, at least portions of the Arabian trade routes could certainly have placed it directly or indirectly, i.e., through its supporters, in CEn Haseva any time from the eighth century BCE onward (Ephal 1984:81-111).

Another central question facing the site's researchers concerns the destruction of the Stratum 5 fortress. When did the fortress cease to function and why? Who was responsible for its collapse? Was it brought to ruin in a fierce military contest between Edom and Judea? Or is there a connection between its destruction and the earthquake mentioned in Amos and Zechariah (Amos 1:1; Zech 14:5)⁷

The book of Amos explicitly records the beginning of his service: "...in the days of Uzziah king of Judah, and in the days of Jeroboam the son of Joash [Jehoash] king of Israel, two years before the earthquake" (Amos 1:1). This earthquake was apparently of such strength that it left a lasting impression on the nation's consciousness, as expressed in the words of Zechariah- "... you shall flee, as you fled from before the earthquake in the days of Uzziah king of Judah (Zech 14:5). Scholars date the quake to ca. 760 BCE and attribute to it the destruction of Stratum VI at Hazor (Yadin 1972:113,179-81), Stratum IV at Lachish (Ussishkin 1977:52), and Gezer (Field XI; Dever 1992:28*-30*). Other sites linked, albeit speculatively, to this earthquake are Deir 'Alla (Phase IX; Ibrahim and Kooij 1979:48), and Beer-sheva (Stratum II; Aharonil1973:107-8). Based on the destruction debris and its configuration, we believe that the quake mentioned in Amos and Zechariah was responsible for the destruction of the Stratum 5 fortress gate complex at (En Haseva. Enlargement of this fortress (to ca. 100x100 m) may have taken place after the earthquake. Final destruction of this stratum may have occurred closer to 735 BCE, during the reign of Ahaz, king of Judea (2 Kgs 16). The biblical record tells us that the Judean kingdom under Ahaz was threatened not only by Syria and Israel, but also by the Edomites and Philistines (2 Chr 28:17- 18; Ahlstrom 1993:684). An Edomite takeover of Elath is placed by the historian of Kings in the time of Ahaz (2 Kgs 16:6).⁸ ...

Although Cohen and Yisrael (1995) suggested that the Iron Age II "Middle Fortress" was most likely damaged by a mid 8th century BCE earthquake mentioned in the Book of Amos (1:1), recent work casts doubt on this suggestion. Cohen and Yisrael (1995) suggested that this fortress, which they dated from ceramics to the 8th-9th centuries BCE, suffered final destruction via either human agency or an earthquake. They dated earthquake damage to ca. 760 BCE relying on historical texts and comparison to archaeoseismic damage at other sites rather than precise archaeological dating from En Haseva and they dated it's final destruction by human agency to ~735 BCE - also based on historical texts.