Ritual Kitchens and Communal Feasting: Excavating the Southeastern Sector of the Ataruz Temple Courtyard, Jordan

Abstract

This paper presents the results of the 2010–23 excavations of the southeastern sector of the Iron II temple at Khirbat Ataruz, Jordan, revealing a purpose-built kitchen complex and a rock-cut libation and animal slaughter feature. Abundant animal bones were collected from the eastern section of the area. Zooarchaeological analysis reveals that faunal remains are dominated by sheep and goats—species central to sacrificial rites—with smaller yet consistent contributions from deer and cattle, and a notable absence of pig bones. Together, these exceptionally well-preserved remains provide rare evidence of integrated cultic architecture, encompassing culinary, sacrificial, and communal feasting, and shed new light on the design of sacred space, religious practices, and temple meals in the southern Levant during the 9th–8th centuries BCE.

1. Introduction

Khirbat Ataruz crowns a ridge east of the Dead Sea, about 24 km south of Madaba and 15 km northwest of Dhiban in modern Jordan (Figure 1). It was a major cultic center in the early to mid-9th century BCE, when a substantial temple was located on top of the acropolis (Ji 2012) (Figure 2). Throughout the late 9th to early 7th centuries BCE, the site functioned as a Moabite settlement before falling into long-term abandonment, only to experience reoccupation in the late Hellenistic, early Roman, and Islamic periods (Ji and Schade 2022). Ataruz also owed its prominence to its position at the junction of routes from the Dead Sea, Wadi Sayl Hadan, Madaba, and Dhiban, which supported its strategic importance and commercial exchange (Ji 2016).

We first offer a concise synthesis of prior discoveries from Ataruz to set the stage for the current paper on the southeastern section of the Iron II temple courtyard. This overview may help readers unfamiliar with previous Ataruz publications appreciate the broader architectural and ritual context against which the 2010–23 excavations of the southeastern sector can be understood (Ji 2012; Ji and Schade 2021, 2022).

Perched in Field A atop the acropolis, the Iron IIA temple complex comprises three parallel halls, as illustrated in Figure 3, where “R” denotes rooms, “I” installations, and “W” walls. The central chamber (RA01) served as the principal sanctuary, its rear wall dominated by a standing cultic stone and a stepped offering platform. Excavations here recovered a rich assemblage of ritual paraphernalia, including bull-embossed storage jars, kernoi, cup-and-saucers, a bronze plaque bearing a serpent motif, finely worked stone mortars, and multiple oil lamps (Ji 2012). To the west lies the Sacred Fire Room (RA04), centered on a square stone altar (IA05). Two raised platforms (IA12 and IA13) in the eastern hall (RA11) suggest the presence of additional specialized cultic functions. Flanking this tripartite structure are stepped “High Places” (IA08 and IA15), further underscoring their ceremonial alignment.

Immediately outside the central temple building lies its inner courtyard, which contains five stone altars (IA17–IA21) and a subsidiary cult chamber, RA15, often referred to as the Bull Statue Room. Within the room, a terracotta bull statue atop an offering table (IA35) embodied the temple’s divine patron; worshippers would have approached this focal point to present offerings and pray (Ji 2012). The chamber also yielded a lion statue and a decanter with four handles, each fashioned as a female figurine. Along the western edge of the courtyard, a stepped seating embankment (IA22) afforded spectators panoramic views of sacrificial burnings at the frontal altars.

Beyond these confines lies the outer courtyard (Fields A and E), where we uncovered the Moabite Sanctuary (RE01) bearing the so-called Ataruz Inscription, a stepped platform (IE02), and a square altar (IE01), discoveries that underscore the site’s potential political and religious connections to the Mesha Inscription and the biblical account in 2 Kings 3 (Bean et al. 2018; Ji 2018). Further east on the slope (Field G), multiple steps—far surpassing mere functional requirements in length—form a monumental stairway that links the city gate below directly to the temple precinct, thereby amplifying both the processional drama and spectator anticipation (Ji and Schade 2021).

Ataruz ranks among the largest and best-preserved Iron II cultic precincts in the southern Levant. Its extensive architecture, stratigraphic sophistication, ritual complexity, and epigraphic association with the Moabite King Mesha render it a cornerstone for interpreting Iron Age religion and cultic practice in ancient Israel and Jordan. This paper concentrates on the southern outer courtyard (Squares A28–A31) and adjacent sector (Squares E6–E7) to assess hitherto mostly unpublished evidence of possible animal slaughter, blood libation, food preparation, and communal dining—features that promise to enrich our understanding of temple ritual and religious life at Khirbat Ataruz. The remainder of this paper is organized into two main sections: the first presents the excavation results; the second contextualizes these findings within a broader archaeological and textual framework, compares them with major Iron Age cultic sites in Israel and Jordan, and discusses their implications for future research.

2. Temple Kitchen

2.1. Room A16

RA16 in Square A29 is a small, rectangular chamber measuring 2.2 × 3.2 m internally. Its perimeter consists of four masonry elements: Wall A48 to the northeast, Wall A49 to the southeast, Wall A54 to the southwest, and the southern segment of Wall A50 to the northwest.

Although all walls are constructed of semi-hewn limestone blocks, their thickness and bonding differ according to the character of the adjoining space. WA54 is 0.8 m thick and built in two block courses; WA48 consists of a single course, 0.5 m thick. WA49 and WA50 each measure 1 m in thickness and comprise two courses. These variations reflect the external functions: beyond WA50 lay the Bull Statue Room, while outside WA49 stretched a bedrock area possibly used for animal slaughter and blood libation. WA54 fronted a principal alley linking the outer and inner courtyards, and WA48 adjoined RA17, a slightly smaller rectangular chamber serving as RA16’s antechamber.

RA16 was excavated in two campaigns—its northwestern half in 2010 and its southeastern half in 2017. During the 2010 season, removal of the overlying topsoil (Locus A29:L1) and the underlying earthy deposit (A29:L9) produced more than two dozen mid-Islamic sherds, together with smaller quantities of early Islamic, late Hellenistic, and Iron II pottery.1 Although no contemporary architectural features were evident, A29:L9 corresponds stratigraphically to A29:L2, which yielded a semicircular stone slab (ca. 30 cm diameter) found horizontally in situ just beneath the surface, adjacent to WA50. A29:L2 represents a mid-Islamic floor or hard surface; this slab may be interpreted as a mid-Islamic installation. A comparable example can be found in Square E8, where two rectangular slabs resting atop a partially collapsed Iron II wall served as the threshold for a mid-Islamic room that repurposed earlier masonry. At Ataruz, mid-Islamic occupants commonly reused Iron II building materials—laying earth-beaten floors within preexisting structures and adding small stone fittings for doors, lamp holders, and grain-processing millstones. These installations typically occur between the topsoil and the Iron II rock-tumble layer. The RA16 slab fits this pattern perfectly, supporting its attribution to the mid-Islamic occupation.

Beneath A29:L1 and A29:L9 lay a 20 cm-thick earth layer (A29:L10) richly mixed with fine ash. This deposit sealed a stone-paved surface (A29:L11) on which four intact Iron II vessels and objects remained in situ: a small storage jar, a large krater, a carinated bowl, and a ball-shaped pestle (Figure 4). The storage jar rested against Wall A48 by the entrance. The krater and bowl lay approximately 1 m and 1.3 m to its northwest; the pestle occupied the space between the jar and the krater. Additionally, scattered across the pavement were numerous broken sherds (mainly bowls, kraters, and storage jars), indicating intensive use. Although later seismic disturbance appears to have disrupted the slabs, several flat blocks confirm the original Iron II stone-paved floor. Beneath that stone pavement, a 15–20 cm earth layer (A29:L14) again yielded abundant ceramics, including 19 diagnostic Iron II fragments, above a thin beaten-earth surface (A29:L15) lying directly on bedrock and virgin soil.

In 2017, we removed the eastern balk of Square A29, exposing an installation (IA33) at the southeast end of RA16, built against the WA49 wall (Figure 5). The feature comprised four stone slabs: a slightly convex basalt slab (30 × 45 cm) at its center, whose northwest edge bore unmistakable signs of breakage. Possibly, the inhabitants deliberately fractured a larger rectangular basalt mortar to fashion a smaller, square slab for horizontal placement inside the installation. To the left of this basalt slab stood a rectangular limestone block (30 × 15 × 20 cm) upright on its long side; when uncovered, its top lay about 10 cm below the basalt slab, likely the result of gradual subsidence into soft soil or seismic disturbance rather than original construction. Flanking these two slabs were two upright rectangular stones: on the left side, a semi-hewn block measuring roughly 30 × 20 × 25 cm; on the right, a similarly sized, uncarved natural slab.

In the southwest corner of RA16, a half-broken Iron II cooking pot remained in situ; its heavily blackened surface and friable walls attest to prolonged use. Adjacent to the pot, against the eastern upright slab of IA33, stood a semicircular basalt fragment with its convex face oriented upward. IA33 spans the central portion of the room’s southeast wall, midway between WA48 and WA54, with the cooking pot and half-moon slab concentrated in its southern quarter. In contrast, the northeast corner of RA16—immediately west of IA33—is free of stone fittings and cooking debris, suggesting it was deliberately left clear for purposes not yet identified.

At the midpoint of WA54, a square, flat limestone block lay flush against the wall. Its precise purpose is uncertain; possibilities include use as a lamp base or as support for vessels or other objects that were intentionally kept off the floor. On the other hand, along WA48, Installation A34 comprised a semicircular array of five small stones encrusted with heavy ash, indicating its use as a hearth for cooking or winter heating. Nearby lay two large basalt mortars and a square block (25 × 25 × 15 cm) bearing a central circular depression (15 cm in diameter and 2–3 cm deep)—probably a socket for a wooden roof-support post. The presence of this pillar base indicates that RA16 was roofed or covered with a canopy.

2.2. Function of Room A16

What was the function of Room A16? All evidence suggests that it served as the temple’s kitchen. An in situ cooking pot in the southwest corner, together with thick ash deposits on both early and later floors, attests to sustained culinary activity. Pestles and mortars recovered alongside intact vessels and abundant food-related sherds further corroborate this interpretation. Also, RA16’s location—immediately outside the temple’s central courtyard, adjacent to a probable animal-slaughtering bedrock installation, and near dense concentrations of faunal remains and pottery linked to temple feasting—underscores that its food preparation was primarily for cultic rituals rather than everyday cooking (see below).

The stratigraphic sequence indicates that RA16 underwent two Iron II phases: an initial beaten-earth floor, followed by a stone-paved surface. All key installations—the in situ cooking pot, IA33, IA34, grain-grinding tools, and the complete vessels—belong to this latter, stone-paved period. Thick ash lenses on the original beaten-earth floor—and within the layers immediately above and below the stone pavement—attest to continuous culinary and food-processing activities throughout both phases. Thus, when second-phase builders upgraded the flooring, they preserved RA16’s function as a space for food preparation, even though evidence of the first-phase layout and installations was removed during the refurbishment. Ceramic assemblages from these contexts date the earlier phase to Iron IIA and the later to early Iron IIB.

Installation A33’s function remains enigmatic. At first glance, the basalt slab appeared to be well-suited for grinding grain, fruit, or meat. However, its modest size and its placement between two upright slabs in a narrow recess would force users to risk striking their heads and arms on the wall behind the installation or on the flanking stones. Two alternative interpretations emerge. First, IA33 may have been intended for occasional, small-scale processing of staples—tasks requiring minimal materials and workspace. Second, more credibly, the carefully carved blocks and their arrangement into a unified installation suggest a miniaturized shrine or offering platform, upon which figurines, ritual objects, or modest offerings of grain and fruit may have been placed in tribute to the temple’s culinary activities.

The decision to repurpose a broken basalt mortar rather than a fresh block adds further intrigue. While practical convenience might explain its reuse, it is also plausible that the fractured mortar—once integral to cultic food preparation in the earlier temple phase—was willfully broken and reconfigured as a potential shrine base, transforming its earlier sacral function into the foundation for a new devotional installation. The first beaten-earth floor was probably pre-Moabite, and the later stone pavement associated with King Mesha’s reign may lend chronological and historical support to this scenario.

Several examples underscore the potential spiritual significance of kitchen spaces. Across Iron II Israel, domestic shrines often fused everyday kitchen wares with cultic objects (Dever 2019, p. 5; see also Albertz and Schmitt 2012, pp. 134–36; Hardin 2010, pp. 124–60): at Jerusalem’s Ophel (Area G), miniature stone altars stood alongside grinding stones, chalices, cooking pots, figurines, and arrowheads; Megiddo’s Cult Corner 2081 paired mortars, pestles, and vessels with incense altars, offering stands, a stone table, and burned grain; and at Halif (Room G8005), storage jars and cooking pots shared space with stone altars, a fenestrated offering stand, and a fragmentary female figurine. Similarly, Israelite houses at Lachish and Beer-Sheba exhibit multiple examples of cultic artifacts integrated into their food preparation and cooking areas (Willet 1999, pp. 134–38, 161, 338). In the Roman period, the lararium—a domestic shrine to the Lares and Penates—was frequently located near the kitchen hearth. Excavations of the Villa of Mysteries at Pompeii uncovered a lararium niche and small altar in the kitchen courtyard, highlighting the close association between the cooking fire and family worship during the Roman era (Foss 1994, pp. 280–99).

2.3. Room A17

For RA17, as in RA16, we began by removing the topsoil (A26:L1) and the underlying soil layer (A26:L2), both of which yielded late Hellenistic, early Roman, early Islamic, and mid-Islamic sherds. Beneath these, A26:L6 revealed an almost room-filling “rock tumble” deposit of small to large stones. This stone-collapse layer—ubiquitous across Ataruz—marks the end of the Iron II occupation at the site and likely resulted from an earthquake dated no earlier than the late 7th to early 6th century BCE. Whether this seismic event precipitated the Iron II settlement’s abandonment or occurred after the site had already been deserted remains uncertain (Ji and Schade 2022).

Beneath the collapsed rockfall, A26:L9 reveals an earth-beaten floor severely disrupted by the collapse. It yielded numerous diagnostic Iron II sherds, a sling-stone–shaped pestle, and a painted terracotta fragment—perhaps part of a cultic object. Two courses of stone blocks were added as buttresses to the original WA50 wall when A26:L9 was laid, to stabilize and repair that wall. A26:L12, beneath A26:L9, represents the earliest beaten-earth floor in RA17. Here, irregular bedrock was levelled with fill soil (A26:L13); four walls were erected directly on this surface to enclose the 1.5 × 3.2 m chamber. The floor was densely strewn with diagnostic Iron II sherds, and a basalt mortar remained in situ. Although two small ash lenses were present, their limited extent suggests they do not indicate significant conflagration or cooking activity within the room.

The entrance to RA17 lay in its northeast corner at the junction of WA47 and WA49. The doorway (A26:L15) measured 1 m wide and featured a threshold of three courses of small, flat stone blocks. Spanning two-thirds of the southeast wall, it created a semi-open space with exterior visibility. At the northwest corner, where WA48 terminates, a second 1 m-wide doorway stood. Its single-course stone threshold led directly into RA16. This was RA16’s only door, so access to the temple kitchen was possible only through RA17. Consequently, RA17 should have functioned as an antechamber, and together the two rooms formed one unified food preparation complex for the temple.

3. Libation/Animal Slaughter Area

3.1. Bedrock Outcrop

Excavation of Square A28 revealed a large natural bedrock outcrop once successive soil and collapse deposits were cleared (Figure 6). Dominating most of the square’s area, this outcrop abuts RA16 and RA17, extends beneath WA48, and underlies the two rooms erected on its western flank. What makes this finding architecturally unique is that, unlike elsewhere inside the temple, where builders covered and leveled irregular bedrock before constructing walls, installations, and floors, the ancient craftsmen intentionally left the bedrock surface exposed here. Such deliberate preservation of in situ bedrock within the temple precinct is exceptionally rare at Ataruz.

For excavations, the surface layers of Square A28, comparable to those in Squares A26 and A29, yielded predominantly mid-Islamic ceramics. Beneath the surface layers, three single-course wall lines emerged (Figure 7). First, Wall A28:L5 runs northeast–southwest for at least 3 m and is roughly 50 cm wide, though its full extent awaits further excavation in Square E5. Second, Wall A28:L8 departs southeast from the southwestern terminus of A28:L5 at roughly a 120-degree angle, stretching 2 m and measuring 30 cm in width. Third, Wall A28:L7 connects A28:L5 and A28:L8 at points approximately 1 m along A28:L5 and 2 m along A28:L8 from their junction; it extends 2.5 m with a thickness of 30–40 cm. Stratigraphic relationships posit that A28:L5 was the primary construction, with A28:L7 and A28:L8 added afterward.

These three walls define a roughly triangular enclosure (A28:L16) of about 1.5 m², perhaps intended for storage or refuse containment. Within this feature, three distinct soil layers were recorded: A28:L6, rich in mid-Islamic sherds; A28:L12, a 20 cm-thick, ceramic-rich layer containing exclusively Iron II sherds; and A28:L17, a 10 cm-thick layer likewise composed solely of Iron II ceramics, many of which dated to Iron IIB, including a Moabite square-rim cooking pot and painted-ware sherds. The lowest soil layer (A28:L17) abuts the basal courses of walls A28:L5, L7, and L8, indicating that A28:L5 and the triangular installation date to the Iron IIB period.

Beneath wall A28:L5, slightly to the east, lies an earlier wall line (WA56) measuring approximately 70 cm thick, 4 m long, and 50 cm high. Unlike A28:L5, this feature is built of two solid courses of semi-hewn limestone blocks, giving it a more robust profile. The wall trends northeast–southwest, running parallel to WA49 at a 5 m separation, and its southwest terminus appears to align with WA49’s end, though that section has yet to be fully excavated. Both walls rest directly on bedrock, with their base elevations differing by ca. 80 cm due to the slope of the underlying rock.

The bedrock outcrop exposed in Square A28 is an irregular, amoeboid mass of light-gray limestone, divided into two principal lobes, one to the southwest and a smaller one to the northeast, by a shallow, trough-like depression. Its surface is somewhat smoothed and gently undulating, following natural bedding planes. Yet, the central “saddle” between the lobes bears three distinct, roughly parallel grooves (each approximately 10–15 cm long, 2–3 cm wide, and gently curved), whose sharp edges stand in subtle relief against the more weathered limestone. Obliquely to the right of these grooves is an oval–teardrop depression, likely a deliberate drainage or utilitarian feature; similar shallow pits and irregular hollows appear on the lower left and mid-slope. Directly upslope of these striations is a circular, cup-shaped hollow about 25 cm in diameter and 20 cm deep, neatly cut into the bedrock. In plan and elevation, the exposed bedrock appears as a low, semi-rounded platform, with its center featuring a combination of linear incisions and a cupule, suggesting localized tool work rather than natural erosion within an otherwise smoothly worn limestone outcrop.

The purpose of this rock feature and its prominent placement within the courtyard remains open to interpretation. Just below the cup-hole’s rim, two broken-knob marks—each roughly 3 cm in diameter—were found about 5 cm beneath the surface. Whether they originally stood as separate projections or were linked by a single bridge ridge is unclear. The cup-hole may have supported the bases of jars or other storage vessels. Yet the presence of potential internal projections invites a more specialized reading: perhaps these knobs (or a ridge) cradled the neck of a sacrificial animal during slaughter. In that scenario, blood would have flowed along the paired grooves into a shallow trench defined by WA56—a modest, field-wall–style dam built to a height of just 50 cm—thereby containing the libation and demarcating the ritual area.

During the 2017 field season, we poured water in stages into the cup-hole to observe its drainage. Small volumes first flowed eastward along the natural cavity in the northeast lobe, but as we increased the volume, the liquid spilled into the central grooves and through the saddle before pooling behind WA56. By analogy, the abundant blood from an animal slaughter—such as a sheep or goat—would have more than filled the cup-hole, overflowing into the eastern lobe and spilling outside its rim toward the central channels. This pattern of flow suggests that both the grooves and the surrounding wall were intentionally designed to guide and contain fluids.2 Although definitive proof remains elusive, the internal knobs within the cup-hole, the channel network, and the low dam-style wall lend support to the interpretation of the bedrock as a setting for ritual slaughter and blood libation—perhaps even an omen-reading ceremony based on the flow patterns.

Two successive floor surfaces (A28:L37 and A28:L33) were documented between WA56 and the base of the bedrock outcrop. The earlier deposit, A28:L37, was laid directly atop the irregular bedrock and virgin terra rosa soil (A28:L40), reaching a thickness of up to 30 cm in places. In contrast, A28:L33, at just 10 cm thick, appears to be an overlay added to repair or consolidate the worn surface of A28:L37. Both strata yielded many sherds, predominantly assignable to Iron IIA rather than Iron IIB–IIC. Since WA56 was constructed on the A28:L40 surface and then sealed by A28:L37, its erection likewise dates to the earlier Iron IIA phase.

During the early Iron IIB period, two new soil deposits were laid above these Iron IIA surfaces, filling the bedrock enclosure up to the crest of WA56 and extending to the base of WA49, thereby capping the exposed bedrock and covering WA56. The lower fill (A28:L26) produced a range of diagnostic Iron II sherds alongside fragments of Moabite painted ware. Above this, another compacted floor surface—spanning loci A28:L20, L24, and L25—yielded an Iron IIB cooking pot and additional Moabite painted wares. The foundation of the triangular installation (A28:L16) rests directly on this new floor, positing that the A28:L5, L7, and L8 walls were built subsequently, likely during the same construction phase. Meanwhile, WA49’s eastern face was reinforced with a single-course addition (A28:L30). Evidence also suggests that the builders partially blocked the east end of the covered bedrock area with an additional wall or stone installation; its original form is now indeterminate due to severe post–Iron II earthquake damage.

In summary, the bedrock outcrop was utilized exclusively during the Iron IIA phase, bounded by WA49 to the northwest and WA56 to the southeast. In the subsequent period, the surface was intentionally buried beneath new deposits and capped with a fresh floor. During this remodeling, WA49 was reinforced with an eastern buttress, and the wall A28:L5—with its adjacent triangular installation—was constructed. Ceramic evidence from the area dates these alterations to the early Iron IIB period.

3.2. Courtyard Alley

An alley in Squares A29 and A30—2 m wide and 8 m long—seemingly served as the main access to the Iron IIA temple’s inner courtyard. It rose gently by about 50 cm from the bedrock area of the outer courtyard and was flanked by two walls: A54 to the north and A55 to the south.

This passage underwent repeated phases of leveling, beaten-soil paving, and modification throughout the Iron II period. First, its irregular bedrock surface (A29:L16) was filled to produce a flattened platform. WA55 was erected atop that surface, and the alley received its initial beaten-soil pavement (A29:L12). The fill and pavement layers produced diagnostic Iron IIA sherds, most likely dating to the 9th century BCE. Later, the alley was resurfaced with a compacted, beaten-soil layer (A29:L6) rich in Iron II sherds and containing a few Moabite painted wares. Immediately afterward, a 2.2 m-long, 50 cm-wide blocking wall (A29:L7) was erected between WA54 and WA55 across the alley, apparently to restrict access to the Iron IIA temple’s inner courtyard. The pottery assemblage dates both the hard surface and the obstruction to the early Iron IIB period. Later, the blocking wall was dismantled and the alley reopened. A new beaten-earth layer (A29:L5) was laid on top of the former pavement to restore the ramp. This final beaten-soil surface, dated to the late 8th and 7th centuries BCE, remained in use until the end of the Iron IIB period.

WA55 marks the southern edge of the alley. It is 1 m wide and built of two to three courses of semi-hewn limestone blocks. The wall runs at least 8 m along a northwest–southeast axis and dates to the 9th century BCE, contemporaneous with the Iron IIA temple. In its northwestern half, WA55 incorporates a built-in bench (IA30) that is roughly 50 cm wide and 4 m long, projecting northward on equally well-dressed blocks. The alley ascends gently. Its narrow, elongated profile presumably evoked a tunnel-like passage that insulated visitors from the outer courtyard’s slaughtering and food-processing activities. This design perhaps fostered a psychological separation, heightening a sense of ascent and underscoring the inner court’s sanctity. Positioned immediately before the inner courtyard entrance and adjacent to the Bull Statue Room, the bench may also have served as a resting spot, a staging area for arranging garments or head coverings, or a waiting place where offerings or sacrificial animals were prepared for placement or immolation on the courtyard altars.

4. Temple Dining Area

As noted, Squares A28 and A29 likely served as the food preparation zone of the Iron IIA temple, wherecooking, processing, and sacrificial slaughter took place. Despite architectural modifications, the Iron IIA kitchen and its associated area remained in use in the early Iron IIB period. Although it remains uncertain whether the soil-filled bedrock area continued to facilitate slaughter, other outer courtyard installations under Moabite control—most notably the RA16 kitchen, the sanctuary in Square E1, and the industrial complex in Square E8—demonstrate clear functional continuity in the Field E area between Iron IIA and IIB. This pattern stands in sharp contrast to Field A’s main temple and inner courtyard, which were systematically sealed and decommissioned during the Moabite takeover (Ji 2012, 2018). If this interpretation is correct, it follows that a nearby space in Field E would likely have been designated for communal dining. We propose Squares E6–E7 as the potential location for this function.

4.1. Squares E6 and E7

Squares E6 and E7 were excavated during the 2017, 2019, and 2022–23 field seasons. Although a comprehensive account of those excavations can be found elsewhere, the following focuses on the features most relevant to our current discussion (Ji and Schade 2021; Ji et al. forthcoming).

A standout element in Square E6 is a rectangular installation (IE07) composed of dressed stones, rising about 1 m above the bedrock. Measuring roughly 0.7 m in width and 1.1 m in length, it is oriented almost exactly east–west at a bearing of approximately 100 degrees. Its squat, pillar-like shape suggests that it was not intended to bear superstructure weight but rather served as a ritual marker delineating a specific sacred area. Comparable pillar-like installations (IA27 and IA32) were previously unearthed in Squares A13 and A27, near the Western and Eastern High Places (Ji 2012). Stratigraphic linkage with the lowest Iron II floor in Square E6 indicates IE07 was erected during the Iron IIA phase.

Immediately south of the WE17 wall, eight successive Iron II floors were exposed in Square E6, including one well-preserved nari-lined surface (E6:L23/L24). The earliest floor, E6:L31 and E6:L38, dates to the early 9th century BCE, whereas the topmost, E6:L15, belongs to the 7th century BCE. Between these lie six intermediate floors, five of which yielded unusually rich assemblages of animal bones, pottery sherds, and tabun fragments.

For Square E7, 2019 excavations revealed a single wall segment (WE22a), projecting 1 m in a northwest–southeast line from the western balk. This feature, about 0.6 m thick and 0.8 m high, is assigned to the Iron IIA period. Further work in 2022 revealed an earlier phase (WE22b) beneath WE22a. This predecessor measured 0.8–1.2 m in width, stood 0.5 m tall, and was founded directly on the bedrock, with WE22a constructed on top of it.

Subsequently, the northern half of Square E7 was excavated, uncovering eight Iron II floors or surfaces of compact beaten earth. The earliest of these combined large and small stones with compacted soil, providing the foundation for the architecture and subsequent floors; the latest horizon lay beneath a collapse of fallen rocks. As in Square E6, the lower Iron II surfaces—particularly those adjacent to WE22—contained dense concentrations of bone debris and ceramics from both Iron IIA and IIB contexts, along with multiple tabun fragments.

For the present paper, the most noteworthy aspect of Square E6 is the exceptional density of animal bones recovered from four Iron IIA and two Iron IIBa floor deposits in its northern section. While faunal fragments appear almost everywhere in the temple precinct, the quantity here is unparalleled. A similar cluster emerged in Square E7’s northwest, where Iron IIA and IIBa layers near WE22 produced abundant animal remains. These patterns suggest that the northern sectors of Squares E6 and E7 probably served as areas for ritual, communal meals, whether following sacrifices, festive observances, or post-worship gatherings, rather than as simple refuse zones. At some cultic sites—such as Dan and Moza—bones were sometimes deposited in pits rather than left on surface floors (Greer 2013, p. 54; Kisilevitz 2015, pp. 150–4). At Ataruz, no such bone pits have been identified yet.

Ritual feasting played a central role in the ancient Israelite temple cult. The Hebrew Bible describes how specific offerings (e.g., the peace offering in Leviticus 3) allocated portions to priests, with the remainder consumed communally as a sacred meal, symbolizing restored fellowship between worshippers, the priesthood, and the deity. Excavations at Tel Dan further demonstrate that such meals were confined to specific precinct areas—the Altar Room in the T-West sector and the monumental altar in T-Center (Davis 2013, pp. 31–89). The substantial bone assemblages in Squares E6 and E7 suggest that sacrificial or ritual feasts likewise occurred in defined spaces at Ataruz, mirroring practices documented at Tel Dan.

4.2. Animal Bones

Square E6 was excavated in its entirety, producing 59 loci; only the northern half of Square E7 was fully opened to bedrock, yielding 33 loci. In total, 3828 diagnostic bones were recovered from these 92 loci. Materials from the partially excavated southern half of Square E7 are excluded from the present analysis.

Here, we report the bone-analysis results for 92 loci in Squares E6 and E7. The implications of these faunal data are discussed separately below. Because ancient bones are fragile and prone to further fragmentation during excavation, washing, drying, or storage, we applied a size threshold: fragments shorter than 1.25 cm (approximately 0.5 in.) were excluded from analysis unless they had sufficient diagnostic features to permit identification of both species and skeletal elements. Consequently, 331 fragments were deemed non-diagnostic, including 118 specimens that—despite exceeding the size threshold—lacked the necessary characteristics for identification.

Table 1. Distribution of Bone Fragments by Field Phase and Animal Species (Squares E6 and E7).

Phase	Chronology	Date	S/G	Deer	Cattle	Bird	Total
1	Surface		20	1	0	0	21
2	Mid-Islamic	13th–15th CE	32	1	0	0	33
3	Rock Tumble	Iron II or Post Iron II	64	0	0	0	64
4	Iron IIBc	L8th–E7th BCE	112	4	0	0	116
5	Iron IIBb	8th BCE	80	4	0	0	84
6	Iron IIBa (TP III)	L9th–E8th BCE	1052	38	13	0	1103
7	Late Iron IIA (TP II)	E–M9th BCE	2139	118	38	1	2296
8	Late Iron IIA (TP I)	E9th BCE	105	5	1	0	111
Total			3604	171	52	1	3828

Note. S/G: Sheep and Goats; TP: Temple Period.

summarizes the results. Sheep and goat bones dominate the assemblage, representing 94.1% (n = 3604) of all identified bone fragments. Bovine remains constitute just 1.4% (n = 52), and only a single bird bone was recovered. Notably, deer bones number 171 (4.5%), more than three times the total of bovine specimens.

Each locus was then assigned to one of eight Field Phases (FP1–FP8), from the surface level (FP1) to late Iron IIA (FP8); their proposed date ranges are also given in Table 1. FP7 corresponds to Temple Period II (early to mid-9th century BCE)—the primary Iron IIA temple phase—and contributes 60.0% (n = 2296) of all bones. FP6, corresponding to Temple Period III (the Moabite sanctuary phase, late 9th–early 8th century BCE), accounts for 28.8% (n = 1103). Together, these two phases comprise 88.8% of the assemblage. Temple Period I (FP8) represents 2.9% (n = 111), Iron IIBb (FP5) 2.2% (n = 84), and Iron IIBc (FP4) contributes 3.03% (n = 116). Bone fragment representation for the mid-Islamic period (FP2) is minimal at 0.9% (n = 33) (Figure 8).

Within Temple Period II, sheep and goats comprise 93.2% (n = 2139) of the assemblage, with deer at 5.2% (n = 118) and cattle at 1.7% (n = 38). Temple Period III shows a similar pattern: sheep and goats 95.4% (n = 1052), deer 3.5% (n = 38), and cattle 1.2% (n = 13). Temple Period I similarly features sheep and goats at 94.6% (n = 105), deer at 4.5% (n = 5), and cattle at 0.9% (n = 1). The dominance of sheep and goats continues in the post-temple Iron IIB phases: in Iron IIBb, they constitute 95.2% (n = 80) and in Iron IIBc, 96.6% (n = 112). Deer account for 4.8% (n = 4) in Iron IIBb and 3.5% (n = 4) in Iron IIBc. Notably, bovine remains are absent from these two late Iron Age IIB phases and the mid-Islamic period.

Two additional points warrant brief comment. First, sheep and goats comprise roughly 95% of the identified bones, which may suggest that ovicaprine meat was consumed 20 times more frequently than that of deer or cattle. However, bone density and fragmentation must be considered: sheep and goat long bones have thinner cortices and lower overall density than those of deer or cattle (Lyman 1994), making them inherently more prone to breakage both during butchery and post-deposition. As a result, ovicaprine remains tend to shatter into many small fragments—each counted as a separate specimen—thereby inflating their apparent abundance. In contrast, the more robust deer and cattle bones survive as larger and fewer fragments and are underrepresented in raw counts. Notwithstanding this bias, it remains clear that sheep and goats were the predominant species sacrificed and consumed at the Ataruz temple.

Our 95% ovicapricous identification rate markedly exceeds the 65% and 82% reported for sheep and goat bones at Mount Ebal and Tel Dan (Hawkins 2012, p. 63; Greer 2013, p. 62). A fuller discussion must await the completion of our analysis of the entire Ataruz assemblage beyond Squares E6 and E7, although these discrepancies may reflect distinct ritual or offering practices at each site. They also could be shaped by each site’s geographic setting, which affected the availability of sacrificial animals: Mount Ebal in the central hill country, Tel Dan in the Galilee region, and Ataruz on the Transjordan Plateau.

Second, of the E6-E7 bones analyzed, three specimens from Temple Period II exhibited clear butchery marks. A vertebra from E7:L28 bears a deep cut on its dorsal surface, while two femora from E6:L29 show three parallel, shallow incisions consistent with slicing. These features are seemingly characteristic of stone-tool use. Chipped flint blades and scrapers—common finds in the Ataruz temple courtyard—were likely the main implements for meat processing. In addition, roughly a dozen fragments from E6:L55 and E7:L22, also dated to Temple Period II, are charred: their surfaces blackened by fire. These include ovicaprine ribs, tibiae, vertebrae, and a single goat horn. Together, the cut marks and charred remains posit that carcasses were both butchered and roasted, while other portions may have been simmered in water near or within the dining area. The abundance of tabun-oven fragments in the area suggests that bread was also baked on-site, supporting a combined model of meat and grain preparation within the temple precinct.

5. Discussion

5.1. Temple Kitchen

Identifying kitchen and cooking areas in ancient Israel and Jordan during the Bronze and Iron Ages requires close attention to archaeological remains and artifact assemblages, since purpose-built “kitchens”—as we understand them today—rarely existed. Instead, food preparation was woven into multifunctional spaces, whether in courtyard niches or the corners of larger chambers. Characteristic structural elements of kitchens include hearths, tabun ovens, stone-lined fire installations, and work surfaces, often tucked into the corners of rooms (Pritchard 1985, p. 72; Shafer-Elliott 2016; Zorn 2023). They are usually paired with grinding implements, such as pestles and slabs, attesting to on-site grain processing. The ceramic assemblage—cooking pots, storage jars, kraters, and bowls—further signals culinary activity, while faunal remains, particularly charred and butchered animal bones, provide direct evidence of food preparation and consumption.

Excavations suggest that in the Bronze Age Levant, cooking was indeed seldom carried out in purpose-built kitchen rooms; instead, hearths, ovens, and grinding implements typically appear in open courtyards or multipurpose areas (Albertz and Schmitt 2012, pp. 64–65; Netzer 1992, p. 196). During the Iron Age, the low, domed clay tabun oven became the standard for baking bread (King and Stager 2001, p. 67; Shafer-Elliott 2016). In Israel’s four-room houses, the central courtyard often served as the kitchen, featuring tabun ovens, grinding installations, and pottery (Faust 2012, p. 215; Ji 1997; Shiloh 1970). Differently, Bronze Age palaces and public buildings often utilized multifunctional rooms, along with adjacent courtyards, which featured larger or multiple ovens, extensive pithoi storage, and richer faunal assemblages, as seen at Megiddo, Hazor, Shechem, and Beth Shean (Ben-Tor 2016; James 1966; Loud 1948; Oren 1992). Larger-scale food processing in Iron Age administrative complexes at Lachish (Ussishkin 2004) and Megiddo (Loud 1948; Finkelstein et al. 2000) parallels the Bronze Age distinction between domestic and elite provisioning (Reich 1992). Nevertheless, finding an indoor space that was distinctly set aside for cooking is still rare even in palaces and public buildings.

Likewise, instances of kitchens within clearly defined temple precincts remain sparse in the ancient Levant. In the north, the temples at ʿAin Dara and Tell Taʿyinat feature architecture reminiscent of the Solomonic temple; however, excavations have not uncovered dedicated kitchen areas within their precincts (Harrison 2012; Novak 2012). Similarly, in the southern Levant, Tel Moza—famous for its monumental temple and courtyard altar—shows no signs of kitchens or food-processing facilities (Kisilevitz 2015). The Judean temple at Arad also lacks evidence of cooking or food preparation inside its sacred area. The same applies to smaller Iron Age cultic structures at Hazor (Ben-Ami 2006), Beth-Shean (Mullins 2012), Lachish (Zukerman 2012), and Yavneh (Kletter et al. 2010).

In Transjordan, excavations at Mudayna ath-Thamad uncovered a small, square cultic room near the city gate, notable for its concentration of cultic artifacts, including altars and incense burners. Yet no evidence indicates that food was served or that communal meals were shared in this space (Steiner and Daviau 2024). The WT-13 site represents a wayside shrine on the northeastern Dhiban Plateau (Daviau 2017). This site yielded many anthropomorphic figurines and statues from a rectangular cultic building, which was furnished with a platform and bench. North of the building, four ovens were discovered with animal bones, ashes, and utilitarian vessels, which suggests that food was prepared and consumed there. However, the cultic building belongs to Stratum II, while the ovens were assigned to Stratum III. In other words, the shrine was erected only after the cooking installations were decommissioned. Excavators associate the communal cooking with food offerings for the dead interredin nearby cave tombs, rather than with cultic activity in the shrine itself.

One potential exception to this general pattern is the Iron II temple at Pella in northwestern Jordan (Bourke 2004; 2013). Excavations south of the site have revealed a multi-phase complex spanning the Middle Bronze Age through Iron II. Importantly, rooms adjacent to and within the Iron II temple structure yielded evidence of food preparation and offerings: burnt bones, large storage jars, ceramic incense burners, and libation vessels. These finds suggest that cooking and food processing were integral to the temple’s cultic functions and may indicate the potential presence of dedicated chambers for preparing sacrificial meals or communal feasts during the Iron II period at Pella.

In addition, as stated, excavations within the Iron II sacred precinct at Dan have revealed vast concentrations of animal bones, providing clear evidence of large-scale communal feasting. Building on Biran’s initial report (1994), Greer (2013, pp. 53–97) and Davis (2013, pp. 42–82) demonstrate that these faunal remains were deposited during structured ritual events, not randomly discarded. The dense assemblages occur in paved courtyards and beside central ceremonial platforms and altars, with Davis’s stratigraphy identifying many deposits as secondary refuse from cleared feast-sites elsewhere in the precinct. Sheep and goats dominate the bone corpus, some of which bear butchery marks indicative of standardized preparation (Wapnish and Hesse 1991). These bone layers intermix with ash and abundant ceramics, linking animal consumption to both food preparation and communal dining (Davis 2013; Greer 2013).

However, despite clear evidence for large-scale ritual food consumption, the excavations at Dan failed to discover a distinct, architecturally defined room or space that served as a centralized kitchen within the sacred precinct. This was also the case for Pella. The analyses of final bone-deposition patterns by Greer (2013) and Davis (2013) suggest that at Dan, ritual meal preparation likely took place in open courtyards or ancillary, multi-functional rooms adjacent to the primary temple spaces.

Two more promising examples come from Horvat Qitmit and Tell Damiyah. Horvat Qitmit, an Iron II desert cultic site in the Negev, comprised two complexes (Beit-Arieh 1995). Complex A featured a three-roomed building, a platform, an altar, and a stone basin, while Complex B included a cultic room with a standing stone and an adjoining courtyard strewn with sherds and animal bones, including fragments of cooking pots. The excavator suggested that this courtyard was used for cooking and food consumption. If so, the standing stone may have marked the place where sacrificial animals were both prepared and eaten in connection with altar rituals. Still, no ovens or cooking installations were uncovered. The evidence overall suggests that food was prepared and consumed in the courtyard, rather than in a designated kitchen or dining hall, although it remains possible that one of the adjoining rooms in Complex B served this purpose.

Tell Damiyah, located near a ford of the Jordan River in the central Jordan Valley, offers another promising case (Petit and Kafafi 2016). Excavations revealed a rectangular mudbrick building identified as an Iron II temple, complete with platforms and human and animal figurines. Immediately south of the sanctuary lay a two-room building containing grinding stones, mortars, large storage jars, and abundant cooking-pot sherds, set on beaten-earth and cobblestone floors. This combination of artifacts may indicate food preparation, and the building’s proximity to the temple suggests that its activities were closely connected to the sanctuary. If this interpretation is correct, the structure may represent a rare example of a kitchen directly associated with a cultic complex.

In sum, there is little doubt that food preparation accompanied ritual practice at Iron Age temples and cultic sites in the Levant (Pardee 2002, p. 226). Worshippers were expected to provide sustenance for their deities. Yet finding clear architectural spaces explicitly designated for cultic food preparation remains mostly elusive. Future excavations at sites such as Dan and Pella may reveal such facilities. For now, among those identified in the reports, Horvat Qitmit and Tell Damiyah stand as the most plausible—albeit incomplete—candidates for cultic precincts with associated food preparation and consumption areas.

In this respect, RA16–17 at Ataruz would stand apart, exemplifying the defining features of an ancient kitchen within a sacred setting. Located in the temple courtyard and adjacent to a libation and slaughter zone during the Iron IIA period, it may embody a deliberate fusion of culinary and animal slaughter functions. Its material assemblage—an in situ, heavily scorched cooking pot; storage jars, bowls, and kraters resting on a floor blanketed by dense ash; a stone-lined hearth; tabun fragments; limestone pestles; and basalt grinding slabs—attests to repeated food preparation. Although these remains derive from an Iron IIBa floor, it is reasonable to infer that the Iron IIA kitchen closely resembled its successor. Overall, RA16–17 would represent a rare and relatively well-preserved indoor kitchen within an Iron Age cultic precinct, a feature previously hinted at on some sites butnever definitively documented. Immediately adjacent lay the locus containing thousands of animal bones.

5.2. Blood Libation Rock

Evidence for ritual libations, as indicated by rock-cut cup-holes and channels, is occasionally attested in the ancient Levant. Excavations at Gezer, for instance, uncovered cup-marks and rock-cut basins adjacent to standing stones (Macalister 1912), which Dever (2014) interpreted as part of a cultic complex used for covenant-renewal ceremonies featuring animal slaughter and communal feasting. At Qeiyafa, Garfinkel et al. (2014, pp. 284–304; 2018, pp. 211–17) uncovered Building D100, dated to the Iron I period, with rock-cut surfaces featuring cup-holes, channels, and basins. The structure is one of three cultic rooms at the site, distinguished by its standing stone, benches, model shrines, libation vessels, and other ritual installations. In addition, Laughlin (1981) described a large stone basin at Dan, situated beside a cultic platform and intended for water or other liquid libations.

Archaeologists differentiate sacrificial cup-marks from domestic or industrial installations, such as grinding pits, through formal, contextual, and functional criteria (Rosen 1991). Sacrificial cup-marks are typically found in ritual settings, adjacent to altars, standing stones, or sacred enclosures. In contrast, domestic mortars are often located near settlement areas, workshops, or food-processing zones. Morphologically, ritual cup-marks tend to be smaller, shallower, and more symmetrical, often with smooth surfaces from repeated liquid libations. Utilitarian mortars are larger, deeper, and show pronounced abrasion from mechanical grinding. A further distinction is the presence of drainage features: sacrificial cup-marks may connect to carved grooves or runnels designed to channel blood or offerings, elements irrelevant to domestic grinding. Finally, associated finds underscore the contrast: ritual contexts frequently yield votive objects, cultic vessels, or animal bones, whereas food remains and grinding tools accompany domestic installations.

The rock outcrop in Square A28 appears to meet most of these criteria and serves as a possible example of the ritual use of natural bedrock. We documented a cup-hole, shallow grooves, potential channels, and a low demarcating wall. Its location beside the Iron IIA temple courtyard and kitchen, together with the absence of domestic debris and the presence of cultic vessel fragments, would indicate the rock’s sacred function.

For textual references, Ugaritic ritual texts (ca. 14th–12th centuries BCE) detail sacrificial procedures that evoke the ritual handling of blood, both as a byproduct of slaughter and as a primary offering (Pardee 2002, pp. 25–72). For example, KTU 1.39, 1.41, and 1.115 describe liquids, libations, altars, the killing of animals, and the burning of selected meat portions, presumably implying the pouring or spreading of blood onto stone surfaces or consecrated grounds. In KTU 1.119, a funerary rite involves the use of liquids, perhaps including blood, as offerings to the deceased; KTU 1.91 describes a sacrificial pit and recounts ceremonies for the gods that involve wine and blood libations (Pardee 2002, p. 272). These passages indicate that blood was probably ritually poured, not merely discarded, and that such libations were associated with dedicated spaces—temple courtyards or sacred rock installations. Smith and Pitard (2009) also note that, in certain ceremonies, blood and wine were applied directly to sacred surfaces without the use of an intermediary altar.

The Hebrew Bible likewise reflects a similar ritual logic, though it is primarily associated with altars. In Genesis 28:18, Jacob pours oil on a standing stone at Bethel, thereby consecrating the rock through a libation in the absence of a constructed altar. Exodus 29:12 directs that a sacrificial bull’s blood be applied to the altar’s horns and then poured at its base; Leviticus 1:5 and 1:11 prescribe dashing it against the altar’s sides. Leviticus 17:11 underscores its unique atoning efficacy: “for the life of the flesh is in the blood.” Such blood was never shed carelessly; it was intentionally employed in divine–human communication (Milgrom 1991, pp. 230–31). Deuteronomy 12:27 also orders that sacrificial blood be poured beside the altar, and 2 Kings 16:14–15 attests to the persistence of these practices even under Ahaz’s cultic reforms. Accordingly, Milgrom (1991, pp. 171–73, 230–34) characterizes the placement, dashing, and pouring of blood as deliberate, regulated acts within a cohesive Israelite ritual system—a view echoed by Levine (1989, pp. 12–17), who emphasizes its treatment according to precise protocols rather than indiscriminate spillage.

In short, Ugaritic and biblical texts converge on a ritual theology in which blood was poured onto sacred surfaces, from altar bases and stone platforms to natural bedrock. These sources may support the interpretation of the Ataruz bedrock outcrop as an intentionally carved libation installation. While wine and oil may also have been applied, two features point specifically to blood rites: a cup-hole flanked by the broken remnants of knobs—suggesting either twin wall projections or a bridge-like element for securing an animal’s neck or head—and a low, dam-like field wall designed to contain blood within the area; if wine and oil were the primary liquids, such a wall would be unnecessary given the smaller volumes required. Together, these features indicate that animal slaughter and blood libation were potentially performed there in tandem, each an integral component of the rite.

5.3. Implications of Faunal Data

Temple Periods II and III yielded 3399 bone fragments—approximately 90 percent of all faunal specimens from Squares E6 and E7. By contrast, the post-temple Iron IIB layers produced only 200 fragments (84 in IIBb and 116 in IIBc). The late Iron IIB occupations at Ataruz reportedly comprise only domestic debris and agricultural installations, with no distinct cultic structures apart from a few household figurines—suggesting that residential reuse had replaced the site’s earlier religious function (Ji and Schade 2022). The Temple Period II–III assemblage is 17 times larger than that of the subsequent phases. Such a stark contrast—and the dense concentration of bones confined to the Period II–III deposits—is unlikely to be coincidental; it links Squares E6 and E7 to communal temple dining when Ataruz served as a cultic center in the 9th and early 8th centuries BCE.

In Squares E6 and E7, Temple Period II is characterized by four successive layers of beaten earth, totaling 40 cm in depth. Bones, pottery sherds, and tabun fragments are distributed relatively evenly through all four layers. The earliest of these surfaces abuts WE07 and WE22, as well as IE07, all of which remained in continuous use throughout Period II. Temple Period III, by contrast, comprises two beaten-earth floors in Square E6 and three comparable layers in Square E7. These surfaces reach a combined depth of approximately 30 cm, and nari pockets appear sporadically in the earliest layer. Together, these observations attest to the intensive, repeated use of the area for temple dining, despite the violent destruction of the Iron IIA temple between Temple Periods II and III. The floors appear to have been routinely repaired and resurfaced to accommodate the accumulation of bone and pottery debris.

Beyond these observations, the faunal data offer further insights. First, communal meals in Squares E6 and E7 are first attested in Temple Period II, after the main Iron IIA temple was constructed atop the acropolis and the Square E6–E7 area was configured as its outer courtyard. Only 111 bone fragments derive from Temple Period I, compared with 2296 pieces from Period II. Ji (2012) contended that Ataruz initially hosted a modest, rectangular cultic structure during Temple Period I, which was later expanded into the more substantial temple complex of Period II. The relatively scant faunal remains from Period I align with the finding that the Square E6–E7 area lacked permanent or prominent buildings during Period I, with only ephemeral wall lines that were subsequently buried or incorporated into the Period II courtyard. In other words, the E6–E7 area was designated for communal meals only after the site was fully developed into a major cultic center with an outer courtyard appended to the temple complex.

Second, as mentioned, communal feasting persisted into Temple Period III under Moabite patronage, following the destruction of the Iron IIA temple and the construction of the Moabite sanctuary in Field E. The 1103 bone fragments from Period III support these sustained communal meals at this location. However, this total represents roughly 50 percent of the assemblage from Temple Period II, suggesting that cultic activity and associated food consumption declined to about half the earlier level. This interpretation also aligns with Ji’s notion (2018) that the Moabite sanctuary in Square E1 was considerably smaller and less grandiose than the Iron IIA temple. Nonetheless, the substantial number of bones—far exceeding what one would expect from a modest-sized sanctuary—implies that ongoing excavations in Field E may yet reveal additional Moabite cultic structures, installations, and deposits dating to King Mesha’s reign and the subsequent decades, potentially more extensive than those uncovered to date.

Third, the abundance of deer bones is noteworthy. A total of 171 deer-bone specimens—approximately 4.5% of the overall assemblage—is three times the number of bovine bones. Notably, 69.0% of these deer bones derive from loci dated Temple Period II, while Temple Period III yields 22 specimens (12.9%). In Temple Period I and post-temple Iron IIB contexts, deer are represented by only four or five bones each. By the mid-Islamic phase, they had virtually disappeared, indicated by a single specimen.

The Hebrew Bible classifies deer as ritually clean and permissible for consumption, yet does not include them among the authorized sacrificial species (Deuteronomy 14:4–5; Leviticus 1:2). Nevertheless, deer bones were recovered from the Period II precinct at Ataruz, likely constructed and used under Omride rule. This finding recalls the faunal data from the cultic structures at Mount Ebal and Tel Dan Area T, where deer remains comprise approximately 10% and 13% of the faunal assemblage, respectively (Horwitz 1986–1987; Wapnish and Hesse 1991; Wapnish et al. 1977). The proportion of deer bones in our assemblage is roughly half that of these two prominent Israelite cultic sites.

Debates sparked by Mount Ebal center on whether venison served as a sacrificial offering or was consumed mainly as meat for communal feasts (Zertal 2018). One position holds that hunters primarily brought deer for banquets; another argues that the substantial deer proportions at Mount Ebal indicate their role as formal offerings. Incidental consumption seems insufficient to explain such high frequencies. The recurring presence of deer bones in Israelite cultic contexts, such as Dan and Ataruz, may also suggest that the biblical prohibition on offering deer could reflect theological refinements emerging in the later Iron IIB–IIC periods, rather than the origins of the practice. A comprehensive treatment of deer in ancient Israelite religion lies beyond the purview of this paper; however, the notable deer assemblage at Ataruz may warrant a fresh appraisal of venison’s role within Iron II Israelite sacrificial rites.

At a minimum, we would suggest that deer were consumed at the Ataruz temple during both Iron IIA and the Moabite period. Venison may have occupied a distinct status in diet—deemed worthy of ritual or communal consumption within the sacred precincts—despite its exclusion from the Israelite roster of official sacrificial species. Apart from its cultic roles, the Ataruz temple could serve as a public venue for various festivals and communal celebrations. The Hebrew Bible records special monthly feasts on each new moon (1 Samuel 20:5, 24–27) and numerous other gatherings marked by shared meals, including Passover (Exodus 12:7–8), the Feast of Unleavened Bread (Leviticus 23:6–7), the Feast of Weeks (Leviticus 23:21), the Feast of Trumpets (Numbers 29:1–2), the Feast of Booths (Leviticus 23:34–35), Purim (Esther 9:27), and the weekly Sabbath (Exodus 16:23). During similar observances, participants could gather at the temple to share venison alongside offerings of sheep, goats, and cattle. Ongoing analyses of the faunal remains from the inner courtyard—where five sacrificial altars have been uncovered—will help determine whether deer were first presented at the altars before being consumed in communal meals.

A few additional observations warrant short comment. First, the relatively high proportion of deer remains in the Temple Period III assemblage accords with Ji’s proposal (2018) that the Moabites inherited much of the Iron IIA temple’s ritual repertoire and cultic accoutrements. This continuity likely extended to communal feasting: the Moabites, like their Iron IIA predecessors, used the courtyard for ritual meals and animal sacrifices, consuming sheep, goats, deer, and cattle as part of their ceremonial diet. Second, cattle warrant attention. They were the largest and most costly sacrificial animals in the ancient Levant; perhaps they were reserved for higher-order rites such as atonement or communal thanksgiving. Their absence in post-temple Iron IIB and mid-Islamic contexts implies that cattle meat retained its luxury status, seldom consumed in non-cultic settings. Moreover, this pattern highlights that the Iron IIA–IIBa temple phases marked Ataruz’s zenith in economic resources—apart from their religious importance—a prosperity that likely waned rapidly after the reign of King Mesha.

Third, the complete absence of pig bones is noteworthy. In Iron Age Israelite contexts, pork consumption was largely avoided, resulting in pig bones being exceedingly rare at settlement sites. The presence of suid remains has thus frequently served as a marker of non-Israelite—particularly Philistine and some Canaanite—occupation (Hesse and Wapnish 1997; Sapir-Hen et al. 2013). For instance, suids are virtually absent from Iron Age Israelite highland and highland-border sites (Horwitz 1986–1987). By contrast, at Philistine settlements, pig frequencies often exceed 5% of the faunal assemblage, while frequencies vary more widely at Canaanite sites (Hesse 1990; Hesse and Wapnish 1997). Our corpus from Temple Periods I–III and the post-temple Iron IIB phase contains no suid specimens, suggesting that pigs were neither sacrificed nor consumed at Ataruz during the pre-Moabite and Moabite eras. This absence in Periods I and II would accord with Mesha’s assertion that the Ataruz region originally belonged to the Gadites and that the city was (re)built under the Omrides. Sapir-Hen and colleagues (2013) reported that during the 8th–7th centuries BCE, pork consumption increased in northern Israel, while it remained rare in the southern Kingdom of Judah. If this view is tenable, the data from Ataruz suggest that this rise in pork consumption in the northern Kingdom of Israel began under the House of Jehu, following the end of Omride rule. Additionally, our data from Temple Period III and Iron IIB contexts indicate that pork likewise did not figure in the Moabite diet. This perspective merits verification through comparative analyses of fauna remains from other Moabite settlements, given the scarcity of such studies at contemporary Moabite sites.

6. Conclusions

Excavations in the southeastern sector of the Iron II temple at Khirbat Ataruz have revealed a deliberately integrated system of animal slaughter, libation, cooking, and communal dining. Central to this system are a possible kitchen complex, a rock-cut installation for libation and animal sacrifice, and a designated ritual dining area. Faunal assemblages are dominated by sheep and goats, with smaller but consistent proportions of deer and cattle, and the absence of pigs—possibly reflecting species chosen for sacrificial offerings, although the precise role of deer warrants further investigation. The persistence of culinary installations, courtyard meals, and likely sacrificial rites from the Iron IIA phase into the Moabite-controlled Iron IIB phase underscores the resilience of these cultic practices amid the political and ethnic transformations of the 9th–8th century BCE. If the views presented here are plausible, Ataruz uniquely compresses the complete ritual cycle—animal slaughter, blood libation, food preparation, ritual meal—into a compact spatial framework, illuminating how sacred space in the southern Levant was organized to coordinate these core sequential rites. The current analyses, however, would be regarded as interim and provisional, grounded in the evidence currently available, and subject to refinement as excavations proceed and new discoveries emerge.