Composite Female Figurines and the Religion of Place: Figurines as Evidence of Commonality or Singularity in Iron IIB-C Southern Levantine Religion?

Abstract

Much attention has been paid to female, pillar-based figurines from Iron II Judah, and the veneration of a major goddess in that territory. Similarly, female figurines throughout the Levant have largely been treated as evidence of goddess-worship, writ large. While the focus on goddesses and fertility has been critiqued by contemporary scholarship, the prevalence of female terracotta figurines remains a productive ground for critical inquiry. There is still no consensus explaining the dissemination of female figurines throughout Levantine states during the Iron IIB-C and how to interpret the similarities and differences among these corpora. Do the similarities that distinguish the Levantine figurines from those of other regions indicate a widespread diffusion of similar praxis across Levantine religion? Do the unique features of figurine design, technology, and deposition that demarcate the corpora of one Levantine state from another provide evidence for a “religion of place” on a more local scale? How should scholars approach iconographic similarities when interpreting the use and function of figurines in different locales? In an attempt to address these questions, this paper uses Levantine composite female terracotta figurines as a test case to explore the way archaeological data both support and impede a geographically contextualized approach to religious praxis.

1. Introduction

One of the most prevalent figurine types across the Iron II southern Levant is the composite female. Composite figurines consist of a three-dimensional pillar base and body and a head constructed separately and then added to the body by means of a clay tang. The bases are either handmade or wheel-made and either solid or hollow, while the faces are molded. Bases may vary in form, width, and curvature. Gestures are always added to the bodies by hand and range from hands on/under attached breasts, to holding objects, to laying out to the sides. The composite female form is found across almost every southern Levantine polity in the Iron IIB-C, as well as in Phoenicia and Cyprus. This distribution provides an opportunity to examine how interpreters might give meaning to regional variation and consistency across iconographic content, artistic style, and technical production processes.

Owing, in large part, to the fame of the composite style from Judah, i.e., the Judean Pillar Figurines, pillar-based female figurines in the southern Levant have received a fair share of attention (Figure 1). At times, this is because scholars have collapsed together the pillar traditions (alongside other figurine forms) from multiple different regions and locations into a single study (e.g., Pritchard 1943; Holland 1975). Even when one region, like Judah, has been the focus of investigation, it is common to include catalogs incorporating the figurine traditions of surrounding regions (e.g., Kletter 1996). Still, these pillar traditions from locations like Ammon, Moab, northern Israel, and Philistia, have been superseded by the attention bestowed on the Judean variety (notable exceptions exist, such as Amr 1980; Daviau 2014, 2017, 2022; Press 2012; see also the essays in Darby and de Hulster 2021).

One of the primary challenges in the interpretation of figurines is which iconographic and technological attributes are indicative of which social behaviors. If a figurine is the end result of (1) choices reflecting embodied processes specific to particular artisans; (2) how workshops curate molds and habituate production processes across a production line; (3) the intended function of the object; and/or (4) possible divine entities which the artisans had in mind, how would modern interpreters know which type of variation indicated which type of decision-making? For example, does a wheel-thrown pillar indicate a different function than a hand-formed pillar? A different supernatural entity entirely? Or merely a different production line or different production instance? Do two molded heads with different hairstyles reflect different entities or simply different implementations of the same entity contingent on which molds were available to an artisan line? And what do we do about gestures and implements held in the hand? Are they sufficient to indicate a different entity, or are they associated with function and setting rather than identity?

When posed in this way, we immediately see the challenges not only in interpreting figurines from one region, but in identifying the social behaviors that account for variations between composite figurines in different regions or even across different sites. For example, if we argue that the differences between molds used in Judah and molds in Ammon indicate different belief systems or even different divine entities, how should we approach two different head molds used in the same city in roughly the same period? How different would the molds need to be to indicate that the artisans have different supernatural agents in mind rather than simply different stylistic implementations?

In the midst of these challenges, how do we then make sense of “regional types” or even the possibility of a regional production style that correlates figurines and territories, let alone polities? Should we begin by first establishing the political or cultural boundary lines of a territory and then review the pillar-based tradition in a given area (e.g., Amr 1980)? Should we begin, instead, by looking at the distribution of figurines with particular attributes and then identify their geographic distribution and cultural horizon (e.g., Kletter 1996, 1999b)?

It is too often the case that these objects, or more properly the modern interpretation of their symbols, are taken as manifestations of the disembodied beliefs of a territory. In the fields of biblical studies and ancient Near Eastern archaeology, scholars have often territorialized belief in a particular way, creating a system in which boundary, ethnicity, polity, and religion appear unified (for review and critique, see Whiting 2007). This consists in associating figurines (or other figural artifacts) with the deities attested in a region’s textual record, disproportionately reflecting the worldview of elite officials and institutions.

In point of fact, at least in the case of the uninscribed and largely ambiguous composite females, there is nothing in the iconography that identifies the figure with any particular deity or divine entity—at least in any way that is clearly discernible to modern interpreters (e.g., Darby 2014, pp. 34–60, 398–99, with literature). So, the assumption that the composite figurines represent major deities of a region’s pantheon may be in error from the start. This does not necessarily imply that the figurines represent humans—only that presupposing an association between the figurines and a region’s political religion may mislead interpretation.

2. Pantheons in the Iron II Levant

Several scholars have argued that Iron Age pantheons were simplified when compared with their Bronze Age forerunners. For example, Hundley (2022, p. 182) compares the 100–250 deities associated with Ugarit alone (depending on how they are counted) with the ca. 10 or fewer known deities associated with Israel and her neighbors in the Iron II (see also Pardee 1988). Both M. S. Smith (2010, pp. 99–103, 128–29) and Sanders (2015) have connected pantheon reduction with political developments in the imperial politics of the period.

At the same time, Handy (1994) has noted the rise of mid-level divine entities during the same time range. How do we make sense of what appear on the surface to be opposite trends? Even if textual and onomastic evidence suggests that Iron II praxis coalesced around a smaller number of major deities, this does not mean that efficacy in the divine realm was limited to those deities alone. Tallay Ornan (2009) and others have noted the plethora of ways divine-ness or supernatural efficacy could be attributed to material objects—like symbols, standards, and intermediaries. This is not limited to the Iron II Levant. In Mesopotamian temples, statues dedicated by worshippers could become the objects of their own veneration practices, as could a wide range of ritual objects, production materials, and celestial bodies (Allen 2015, pp. 55–66; Pongratz-Leisten 2011; Porter 2000, pp. 243, 246, 2009; Hundley 2013; Selz 1997).

In sum, the predisposition toward identifying figurines with national-level deities or deities known only from the textual record masks the complex range of options that might actually be associated with these figures. It also makes the national level religious expression reflected in inscriptions the adjudicator of the religion of an entire region, rather than reflecting the diversity of possible religious praxis and practitioners. What happens if we approach composite females as ritual objects attributed with supernatural efficacy rather than representations of major deities designed primarily or solely to receive worship or veneration?

3. A Religion of Place

Regardless of whether we are discussing a national god associated with political economy or a deified object, a case could be made that these elements have geographic purview. In much the same way that Ashur or Yahweh are associated, at least at some level, with territory, is it possible that divine servants, defied standards, and other objects imbued with divine efficacy are also geographically contingent?

Allen (2015, pp. 198–236, 272, with literature) demonstrates the way pantheons across the ancient Near East were deeply embedded in various geographies. Deity names included epithets that identified their associations with nations, cities, and even geographic features, and these geographic determiners often communicate specific and, at times, separate divine entities. The salient point is that most ancient Near Eastern populations likely thought of divine entities and ritual landscapes as associated with particular spaces, places, territories, and/or geographies.

Whenever we discuss regional religion, there is always an implied definition of place. For example, if we were to focus on the “religion of Judah,” whose understanding are we trying to reconstruct? Are we interested primarily in the deities venerated by the majority of people in a location? The deities venerated by the people who were born in a location? Any deities venerated by anyone who happens to live in or travel to a location? The deities venerated by the elite of a specific state? By ignoring these questions in our reconstructions, we betray an understanding of religion that is static, as if belief is tied to territory without regard for variation, temporal development, or contestation. And we default to the stability of institutions—thus, the deities of temples and royal houses, which are not only more likely to survive in the textual record but are given the dominant position associated with state-level sponsorship.

As J. Z. Smith (1978) noted so many years ago, “map is not territory.” Humans create and give meaning to space and place in multifaceted ways. Rüpke and Urciuoli (2023) argue that place or space, as an act of geography-making created by actors and bodies, can even stand in contrast with the attributes of a structure’s actual physical location. For example, urbanity can be reconstituted in the hinterland, such as a country villa built in the format of a city palace. This engagement with social geography has a long history in scholarship. Olson et al. (2013) provide a helpful summary of geographic approaches to religion, arguing that the best analytical pathway is to focus on the co-creation of both “religion” and “space.”

In his summary of the field, Werlen (2021) provides not only a review of scholarly approaches but a systematic discussion of geography-making at multiple societal levels. Focusing our analysis on space-making requires us to pay more attention to specific actors and the way they spatialize their lives and landscapes. Furthermore, Rüpke and Urciuoli (2023) suggest that a careful attentiveness to actors and meaning-making may allow us to view, however dimly, geography-making outside the frame of dominant narratives (e.g., as in elite religion). So, what do we learn if we approach figurine-makers as geography-makers in whose work space and religion are co-constituted?

4. Why Figurines Are a Good Fit

By escaping any easy association with prestige practices, figurines provide an opportunity to view place-making in a different light. Figurine manufacture is largely unassociated with major temples in the southern Levant. Although known Iron IIB-C temple sites are few in number, freestanding figurines are not generally prominent parts of temple assemblages (Daviau 2012; Nakhai 2015). Even shrines only rarely include figurines. For example, in the Israeli southern Negev, where two known shrines have been excavated, neither ‘En Ḥaẓeva (Ben-Arieh 2011) nor Ḥorvat Qitmit (Beck 1995, p. 43) had more than a very small number of unattached freestanding figurines (WT-13 from Wadi ath-Thamad, discussed below, stands in some contrast with this pattern). In comparison, roughly contemporary sites in the same region, like Tel ‘Ira (Kletter 1999b), Malḥata (Kletter 2015), and ‘Aroer (Thareani 2011), all produced figurines, coming from contexts outside shrines or temples. The same pattern is observed in the rest of Judah, where the vast majority of figurines are located in neighborhoods, houses, construction debris, and fills rather than in sacred space. Even at the Arad temple, the number of figurines possibly associated with the temple (at least indicated by the current state of publication) is dwarfed by the number in other contexts (Darby 2014, p. 257 with associated literature).

The makers of figurines are most likely local artisans, and figurines are generally produced in the regions where they are distributed. Although more provenience studies would be ideal, the studies currently completed indicate that figurines are largely (though not entirely) a regional and local affair, created in the locations where they are used. The results have produced no substantive evidence for the organization of elite-sponsored and centralized workshops. (e.g., Kletter 1999b, p. 384; Peterson-Solimany and Kletter 2009, p. 116; Darby 2014, pp. 183–212; Ben-Shlomo and Darby 2014; Ben-Shlomo and McCormick 2021; Gunneweg and Balla 2017).

In fact, the materials out of which the figurines are made are not prestige items controlled or coveted by political elites. The ancient Near Eastern textual record, alongside studies of object construction and distribution, suggest that elite control focused on high status objects, like metals, stones, and fumigants, rather than objects made from clay. This may be because clay sources were readily available or because producers making pottery and other clay objects largely controlled their own production industries (Darby 2014, p. 358, with literature).

Finally, figurine makers have a variety of options available at their disposal, even within a relatively constrained field. One might argue that molds, as on composite-style heads, diminish artisan design control; but the choice and implementation of molds, as well as hand-finishing, indicate both intention and habit on the part of an artisan community, as do the molds’ preservation. Furthermore, on all examples, gestures and implements are hand-modeled, providing further avenues for design choices. And though paint is poorly preserved, this medium might provide the best opportunity of all (McCormick 2023). In sum, figurines were generated through actors, whether influenced by individual choice, ritual need, material requirements, or artistic tradition. If we are correct to hypothesize that ritual implements, like figurines, are attributed with supernatural efficacy and are also associated with place and territory, then the act of producing a figurine in a particular style or with certain content is an act of individual or collective place-making on the part of an artisan community.

5. The Challenges

Several difficulties impact a comparative study of composite figurines. First, the fragmentary state of figurines when excavated means that a body fragment found without its head (a common occurrence) is not indicative of composite form except where a clay tang is found still adhering to the neck or torso, or the body fragment is hollow. This relegates our comments primarily to head fragments in the round that include molded faces (as opposed to heads from flat-lying, fully frontally molded figurines) and examples where heads are still joined with bodies or body fragments (which are few and far between).

The antiquities market and incomplete publication are even greater challenges. Figurines with no known provenience must be excluded from these types of investigation. Figurines with partial context information, such as site-only provenience, provide very little assistance in questions of archaeological distribution and use or discard. These same factors impede a complete understanding of the figurines’ chronological horizon. Incomplete publication, early excavation techniques, and objects acquired on the market make it very difficult to track the rise and spread of the composite technique. Dating is also impacted by the vagaries of structural preservation, stratigraphic sequence, and depositional history.

An additional challenge is the lack of provenience studies for figurines across the southern Levant. Based on the current petrographic studies, it is generally assumed that figurines are largely made in the region where they are found—often in the city where they are excavated. Still, some data suggest that figurines could have come from other regions, like the small number of composite figurine fragments in Jerusalem made from clays originating in the coastal plain (e.g., Figure 2(15); Darby 2014, pp. 158, 461, Table 37) or the small group of composite figurines at Nasbeh that may have been made of Jerusalem rendzina clay (Ben-Shlomo and McCormick 2021).1)

Where provenience studies are lacking, we must keep in mind that a composite figurine found in one location could have been made in a different region. The Jerusalem study suggests tentatively that variation in clay type might correspond to variation in iconographic content (e.g., Darby 2014, pp. 203–10, though this may be more recognizable in pinched head figurines), but it is almost impossible to assume the opposite—that variation in iconographic content necessarily implies that a figurine was made elsewhere (e.g., Darby 2014, pp. 200–2).

6. The Corpora

Despite these challenges, the corpus of composite female figurines is sufficiently large to test whether or to what extent an approach from the geography of religion might provide new or additional insight into the way figurine production and space-making overlapped in the Iron II Levant. Furthermore, by focusing on the composite females in particular, we can better compare issues arising between and across Levantine regions, which might otherwise fade if analysis includes multiple types and styles of figurines all together. While it is true that figurine makers produced many types of figurines in the Iron IIB-C, both anthropomorphic and zoomorphic (e.g., Briffa 2017; Darby 2014; Kletter 1996; Amr 1980; Holland 1975), the issue under consideration here is whether we can see in one particular technological style evidence of regional adaptation, distinction, and/or interaction.

6.1. Judah

The composite style in Judah lasts primarily from the eighth through the sixth centuries BCE, with the possibility of late ninth through early eighth century exemplars (see below). As is well-documented, Judahite molded heads contain almond-shaped eyes, archaic smiles, and between one and six horizontal rows of curls framing the face and covering the ears. Stylistically, the faces can range from more “Egyptianizing,” with thinner, pointed eyes, to rounder features (Kletter 1996; Darby 2014; McCormick 2023). The heads are joined to largely handmade pillar bodies through a clay tang. Neck length varies. While the majority of bodies are solid, several wheel-made, hollow examples also exist, especially earlier in the temporal horizon (Darby 2014, pp. 216–18, 255–56). While there are some bodies that hold objects, like drums or children (e.g., Gilbert-Peretz 1996, p. 114, Figure 12:14–15), the overwhelming majority of specimens have the hands at or under the breasts. An alternate pillar-based format with hand-pinched head is also well represented in Judah, particularly prominent in the Judean Hill Country and in the seventh–sixth centuries BCE (Darby 2014, pp. 252–58), but those are not in the composite style discussed here.

Most examples of Judahite composite figurines are found in the known territory of Judah itself (Kletter 1996, 1999a), with very few exemplars outside the kingdom. A small number were found in Philistine sites on the borders of Judah (Kletter 1996, p. 95; Darby 2014, pp. 314, 328; Press 2012, p. 167) and one was uncovered at Khirbat al-Mudayna ath-thamad in Moab; it was modified by adding a hand-formed Hathor-style headdress (Daviau 2014, p. 3; Daviau 2022, p. 257 Figure 3:5). While individual sites in Judah have revealed a range of head molds, the overwhelming majority depict some version of the short, Egyptian Old Kingdom style headdress with curls covering the ears (Darby 2014, pp. 339–42). This is one reason Kletter (1996, 1999a) concluded that the JPF style group seems to correspond with the territory of Judah.

6.2. Philistia

The composite style in Philistia is marked by a wide variety of head and hair styles. Heads are joined to the body in the same manner as Judahite styles, though body types vary. In some cases, they may appear as more rectilinear (e.g., Press 2012, p. 60 no. 37 from Ashkelon), while in others they learn toward Phoenicianizing bell-shaped and hollow examples (Ben-Shlomo 2021, Figure 5.6:7 from Tell Jemmeh). Faces can be very Egyptianizing (e.g., at Ashkelon, Press 2012, p. 58 nos. 33, 34; p. 60 no. 37; p. 62 no. 42; p. 64 no. 46), though eyes can also be wider (e.g., ibid., p. 39 no. 36; p. 63 no. 43; p. 66 no. 49; p. 69 no. 54). At Ashkelon, some heads include simple, shoulder length hair or veils, either covering the ears (e.g., ibid., p. 58 no. 33) or more commonly with ears showing (e.g., ibid., p. 58 no. 34; p. 59 nos. 35, 36; p. 60 nos. 37, 38; p. 61 nos. 39, 40; p. 62 nos. 41, 42). In some cases, the veil has been added or enhanced by hand (e.g., ibid., p. 59 no. 35; p. 63 no. 44; p. 65 no. 48). The site also produced pointed veils (e.g., ibid., p. 64 no. 45; p. 67 no. 52).

At other sites, like Ekron, some figurines have a more Phoenicianizing appearance, including a hollow, bell-shaped body and molded head with shoulder-length sidelocks twisted in braids (Ben-Shlomo 2021, Figure 5.6:6). This particular fragment from Ekron is significant because it comes from Temple 650 and the first half of the eighth century (Ben-Shlomo 2021, p. 133; Press 2012, p. 211; cf. with Nunn 2021, Figures 4.6a, 4.6b), perhaps indicating the composite style in Philistia spread from southern Phoenicia. This hypothesis may be confirmed by the incorporation of a molded head on one of the rectilinear cult boxes from the Philistine favissa at Yavneh. In this case, petrographic analysis suggests that the object was not created from local clays and may have been generated in the north (Darby 2014, pp. 343–44; Ben-Shlomo and Gorzalczany 2010, p. 152).

Gestures vary much more broadly in the Philistine corpus than in Judah, including holding the breasts, discs, and children, and torsos with arms upraised (Press 2012, p. 168). The Philistine styles are rarely found outside of the region. While the heads have more in common with the composite figurines of Israel and Moab than they do with Judah, the bodies include the breast-holding gesture (e.g., Press 2012, p. 73, no. 63; Ben-Shlomo 2021, Figure 5.6:7) more frequently than do examples from other non-Judahite locations. Provenience for Philistine composite females is mixed, with a large number coming from discard or fill loci. They likely rise in the eighth century BCE, with many if not most examples from seventh century contexts (Press 2012, pp. 199, 205–6; Ben-Shlomo 2021).

6.3. Transjordan

No composite female figurines are known from southern Jordan, but they have been found in northern and central Jordan. Owing to the state of publication for many Transjordanian sites and their stratigraphic sequence, it is difficult to say more than that the composite style arises at various locations during the Iron IIB-C (Amr 1980, pp. 22–35; Daviau 2017, pp. 5, 20). Daviau (2017, p. 5) hypothesizes that Jordanian figurines were kick-started in the Iron II due to Phoenician influence (see also Gubel 1991, pp. 132–36; Markoe 2000, pp. 158–59).

The head styles and gestures for composite figurines found in northern and central Jordan are far more varied than what is found within the boundaries of Judah (e.g., Kletter 1996, p. 92, Figure 10:1–7; Daviau 2014, p. 3). Furthermore, hair styles in Ammon sometimes correlate more closely with varieties found in Israel and the Jordan Valley. Many of these tend to have elaborate hairstyles, head pieces, and jewelry. Discussing inter-regional mold links among plaque figurines (as opposed to composite figurines), Hunziker-Rodewald (2021, pp. 242–43) argues that multiple links connect Ammon, the Jordan Valley, and Israel, while Moab appears to have a more cohesive style group that demonstrates little connection outside of its region. The same might be said for the composite female tradition.

Overall, composite female figurines in Jordanian regions are a less internally coherent style group than the Judahite variety. When combined across Jordan, composite females are outnumbered by frontally molded figurines (Hunziker-Rodewald 2021, p. 224). At individual sites, however, the composite style can be relatively well-represented, such as at Khirbat al-Mudaynah (Daviau 2022, p. 255) and the shrine site WT-13 (Daviau 2017, p. 130). A number of head molds have been preserved from Jordan, including three from Amman, three from Tall al-‘Umayri, and one from Tawaylan (Hunziker-Rodewald 2021, p. 226, Table 10.1).

Jordanian composite molded heads sometimes contain a simple veil (more typical in Moab) but may also include more elaborate headpieces with horizontal or vertical curls. Some can be quite ornate, as illustrated by a head mold from Amman that includes multiple locks of vertical curls to the sides of the face, as well as an ornate headpiece and earrings, comparable to the women in the window motif known from Assyrian ivories (Hunziker-Rodewald 2021, p. 234, Figure 10.4; cf. an example with ornate style head from Press 2012, p. 64 no. 46 from Ashkelon). The length of the headdress varies. The molding of the eyes tends to be wider than the average examples excavated in Judah or Philistia, and most heads depict the ears. The majority of bodies hold objects (particularly drums) or fold their hands on the upper chest (Daviau 2022, p. 255). Bodies and bases are formed by a variety of techniques, ranging from wheel-turned to hand-formed (coiled) with a concave base (Daviau 2014, p. 3; e.g., Daviau 2017, pp. 102–3, WT 479; Daviau 2022, p. 255).

A very large portion of the known total figurine corpus from Jordan comes from the shrine at Site 13 in Wadi at-Thamad (Daviau 2014, p. 5, Chart 1), particularly Stratum II, dated to the eighth–seventh centuries (Daviau 2017, p. 72). Daviau (2017, p. 5) notes that before WT-13 was excavated, the dominant context types for figurines in Jordan were domestic contexts followed by funerary contexts, making WT-13 unusual. It is also one of the only Iron Age shrine sites in the southern Levant with substantial numbers of freestanding figurines. None of the composite females from WT-13 hold their breasts; all have hands at the waist parallel to the body, hold a disc, or play a drum perpendicular to the body (Daviau 2017, p. 99), and the site produced ca. 43 fragments of the composite figurine variety (ibid., p. 130; note there are only ca. 28 actually discussed in the catalog: ibid., pp. 99–106) out of ca. 91–92 total anthropomorphic figurine fragments overall (including attached figures, but excluding larger statues) (ibid., pp. 106, 130).

6.4. Israel

Composite figurines in Israel are also plagued by issues arising from early excavations at sites like Megiddo and Beth Shean and the corresponding publication details (or lack thereof). Published archaeological provenience makes it hard not only to date the specimens but to say anything meaningful about use and discard contexts. Although several examples were found in rooms originally dated to the Iron IIA by early excavators, upon further examination many of these contexts were mixed or are located in rooms and building phases that have been subsequently re-dated to the Iron IIB or later periods (Darby 2014, pp. 314–19 with literature). More recently, two possible composite style heads from Tel Reḥov (with stylistic comparanda from Deir ‘Alla, Megiddo, Tel Jemmeh, and Bethsaida) found in clear stratigraphic contexts suggest that separately molded heads may have arisen as early as the tenth century at northern sites (Saarelainen and Kletter 2020, p. 550 Figure 34.3:21–22, 568).2

Head styles show a much larger degree of variety in Israel, as do the range of gestures on pillar bodies (Kletter 1996, p. 89, Figure 7). Heads may be adorned with a variety of headdresses, ranging from simple horizontal rows to vertical curls, to a mixture of the two. Heads sometimes have a molded headband, diadem, or jewel-piece resting on the forehead. The length of the hair and/or headpieces also vary, from the top of the neck to below the shoulder. Ears are generally visible, and jewelry is sometimes included in the molded design. Facial features vary, with several examples having rounder eyes or eyes pointing down at their outward edges. As above, Israelite sites have closer similarities with other northern and coastal sites and with heads in northern and central Jordan (e.g., two heads from Tel Reḥov in Saarelainen and Kletter 2020, pp. 567–68, Nos. 21–22 on p. 550 Figure 34:.3:21–22). Northern Israelite sites have also produced heads in Phoenician style (e.g., Kletter 1996, p. 91, Figure 9:6).

Northern Israel may also have generated some evidence for one type of transition from fully molded plaques to free-standing figures. Figurine 15 from Tel Reḥov (Saarelainen and Kletter 2020, p. 547, no. 15, Photo 35.15, p. 567) consists of a frontally molded female with arms at the sides but with shortened and flattened legs and additional clay that enables the figurine to stand unaided. The figurine comes from a striated earth layer (consisting of red clay, gray ash, white lime, and brick debris) inside a partially excavated structure in the western side of Area C, dated to the tenth century BCE. Given the presence of grinding and pounding stones, the excavators consider this a domestic space (Panitz-Cohen and Mazar 2020, pp. 23, 39–40). The figurine in question has all the characteristics of a molded female, with the addition of rounded clay on the head, and perforation marks where curls might be indicated. Although the figurine appears in the image to be missing ears, the description makes clear that ears are present on the mold, but they are obscured by the additional hand modeling on the hairstyle. (Saarelainen and Kletter 2020, p. 547). This fragment may represent an alternative method to wheel-formed bases for adapting a figurine to a free-standing position (see also the comparanda from Megiddo, Hazar, Amman, and Tel Jawa noted in Saarelainen and Kletter 2020, p. 567).

Furthermore, the head of Figurine 15 shares certain features with one of the site’s two possible composite heads, Figurine 22 (Saarelainen and Kletter 2020, p. 551, no. 22, Photo 34.22). While not identical, both heads use perforations on the headdress and both include molded necklaces. Figurine 22 was also found in a clear tenth century context, in occupation debris overlying the courtyard of a large courtyard house (possibly associated with the site’s scribes or other elites), alongside a zoomorphic fragment, an amulet, and two inscribed sherds (Panitz-Cohen and Mazar 2020, pp. 23, 47, 49, 52). It is tempting to postulate that both Figurines 15 and 22 provide evidence for the way artisans at Tel Reḥov in the Iron IIA began to experiment with modifications to adapt fully molded figurines for new ritual demands, as well as the way different technological styles might share similar iconographic features.

6.5. Phoenicia and Cyprus

Because it is possible that the composite figurine tradition in the southern Levant was influenced by figurine styles in Phoenicia and Cyprus, some consideration of that tradition is in order. In Phoenician composite females, the bodies are most frequently wheel-formed, bell-shaped, and hold objects, typically a drum or a bird (Nunn 2021, pp. 73–76; Figure 4.6a–b, 4.7). The shape of the pillar has led interpreters to argue that these figures wear a dress (Nunn 2021, pp. 71–72; Press 2012, pp. 204–6). Some Phoenician styles tend to have unique additional hand-formed hair, particularly long side locks or twisted locks added to molded faces (e.g., Kletter 1996, p. 91, Figure 9:2–3; Nunn 2021, pp. 76–77 Figures 4.7, 4.8). Others seem to have a simple headdress (Kletter 1996, p. 91 Figure 9: 5, 8). When the headdress is molded, it may consist of side locks twisted into braids (e.g., Nunn 2021, pp. 73, 75, Figure 4:6b).

The dates for the beginning of the composite style in Phoenicia rest on a small number of fragments; but, based on the available data, the style may arise as early as the eighth century with continuation into the fifth century (Press 2012, p. 172; Nunn 2021, p. 75). Nunn (2021, p. 75) suggests that they may constitute a “south Phoenician tradition,” and notes fragments are found in cultic, burial, and domestic contexts, though cultic contexts are disproportionately well-represented in the excavation record, perhaps skewing our picture of figurine distribution (ibid., pp. 110, 111).

In Cyprus, bodies often take the form of tall and thin pillars holding objects, and the heads are frequently hand-formed, which may nod to the island’s earlier Mycenaean-style pillar figurine tradition. Where female molded heads are present (e.g., the Kamelarga Group, Caubet 2021, p. 55, Figure 3.2a–c), they tend to contain shorter headdresses that show the ears. Gestures vary, including upraised hands, holding infants, and holding a range of objects, like vessels, birds, cakes, and instruments (Averett 2021, p. 313). This style likely arose on Cyprus in the eighth century (with an alternative dating in the seventh century), giving rise to a debate about whether Cyprus or Phoenicia formed the impetus for the composite females (e.g., Karageorghis 1991, p. 13; Caubet 2021, p. 61; Nunn 2021, pp. 77–79; Averett 2021, pp. 307, 311). The style continues into the sixth century (e.g., Averett 2021, p. 312, Figure 13.10, from Larnaca).

7. Discussion: Interpreting Coherence and Diversity

7.1. Iconographic Consistency and Variation

One of the primary challenges in comparing composite forms across regions is to determine how much and which variation matters and in what ways. To review, in the realm of iconographic consistency versus variation, all of these regions share the inclusion of eyes and mouths but vary in the styles of headdress, the presence of jewelry, and the depictions of ears. Mouths are generally closed and smiling. This suggests that, whatever differences characterized each region, the foreword-staring molded eyes and closed, smiling mouths were likely central to the function of all composite female types.

Between regions, sites, and across the same site, eye shape may vary in width and whether the external corners of the eyes point up, down, or remain central. Given that eye shape can range even at the same site (e.g., on Judah, see McCormick 2023, pp. 52–53, 62), it is difficult to associate shape and orientation with regional styles rather than the molds represented at a site or in a production line. As per below, variations in the intricacy of hairstyle and adornment might be more indicative of regional differences.

Gestures range broadly, with the hand-on-breast gesture far more prominent at Judahite sites than in any other region. Most regions have produced evidence for a variety of gestures, even at the same site. While a small number of composite females hold a drum or baby in Judahite territories, they pale in comparison to the number holding drums elsewhere. Holding a variety of instruments and offerings and gestures with hands out to the sides are also far more common outside Judah.

7.2. Technological Consistency and Variation

7.2.1. The Pillar Bases

All of these regions share the production style of a molded head attached to a separately formed body. The evidence suggests that they were painted (though details remain ambiguous), and gestures were added by hand (rather than molded on the body). Clearly, in all regions the figurine had to be able to stand unaided, rather than being leaned against or affixed to structures. From one perspective, harkening to the role female protomes play in lieu of columns in the shrine box tradition (Darby 2014, pp. 321, 332, 346–47), perhaps the makers intentionally incorporated columns as a significant aspect of figurine iconography (McCormick 2023, pp. 164–96). However, a composite female/column could be easily depicted in two dimensions. Functionally, then, whatever else the bases represent, the three dimensional form suggests that figurines needed to be able to stand largely unaided, meaning that the form of the body is not necessarily motivated by iconographic depiction alone but by the actual uses to which the figurines could be put. A truly free-standing figurine has the benefit of moving from place to place as ritual requirement demands.

The variation in approach to the formation of bodies, whether wheel-made or hand-formed, seems to suggest that the way a production line produced the freestanding base was a reflection of style and artisan culture more than of regional homogeneity. Within a given site, like Jerusalem, one might observe a mixture of these technological approaches, though wheel-formed examples in this case might be slightly earlier in the archaeological record (Darby 2014, pp. 215–18). Given the wide variety of base formation techniques and outcomes (some thinner, some wider, some more rectilinear) within each of the various regions, it seems likely that the way a base is manufactured does not reflect regionally-coherent beliefs but the production preferences and/or customs of specific potters or production lines (contra Press 2012, pp. 204–6).

It is worth considering, however, whether the manufacture of hollow pillar bases echoes object function. For example, wheel-turned or hand-coiled hollow bases have similarities in production process to other objects in cultic assemblages that incorporate these same potting methods. If we consider hollow composite females alongside other objects that incorporate vessel-forming techniques, like anthropomorphic and zoomorphic vessels and cylindrical cult stands, then perhaps the manufacture tradition of the hollow base also indicates a kinship between the composite female production process (especially in the eighth century) and the production of mixed form cultic vessels. Many such composite females also hold implements, offerings, and instruments, perhaps suggesting that their makers produced them with the same production habits used to create other types of ceramic shrine equipment and considered them akin to cultic attendants.

7.2.2. The Molded Faces

Figurine makers across these regions maintained head molds throughout the Iron IIB-C. One could argue that it would be easier to make a hand-pinched head in tandem with the body, as is seen in a large number of female pillar-based figurines in Judah (Darby 2014), the north Syrian pillar tradition (Pruss 2021, pp. 344–52), many Cypriot examples (e.g., Averett 2021, pp. 304–14), as well as in Mycenaean figurines and Philistine figurines in the Iron I-IIA (Ben-Shlomo 2021, pp. 121–30; Press 2012, pp. 152–65). In fact, fully hand-made figurines are found throughout these regions even in the Iron IIB-C, though they are more typical for figurines gendered as male or of indeterminate gender. So, what purpose could there be to continue the more laborious process of separately molded heads?

One argument is that artisans are creating stylistic and technological continuity with the older plaque and frontally molded traditions. By establishing the antiquity of molds and continuity with objects like the Late Bronze plaque figurines, the tenth–ninth century molded female drummer tradition (Paz 2007), and the molded females applied to cult stands, altars, and shrine boxes, the practice invests the composite figurines with efficacy and authority. This might especially characterize production lines from sites that also enjoyed a very active and ongoing Iron IIA-C frontally molded figurine tradition (which is true of most Iron II Levantine regions, though much less common in Judah).

For example, at Ashkelon, some heads from frontally molded figurines and some heads from composite figurines share features, like the shoulder-length simple veil (cf. frontally molded figurine head Press 2012, p. 74 n. 65 with heads above). In northern Israelite sites, figurine heads attributed to the pillar type sometimes share similarities with fully molded varieties, (e.g., both from Megiddo: a pillar-based holding drum in Kletter 1996, p. 89, Figure 7:1, 2, 3, 8 and a frontally molded style holding a drum in Paz 2007, p. 36, Figure 2.1:10). Similarities between the head styles of frontally molded and pillar-based figurines can also be found at Tel Jemmeh (Ben-Shlomo et al. 2014, pp. 809–10 Figure 17.3: a–d).

In some cases, the head molds used on composite figurines seem to differ from those used on fully frontally molded figurines in the same period. For example at WT-13, when compared with preserved molded heads from the site’s frontally molded figures (Daviau 2017, p. 93 Figure 4.3), the composite heads have a simpler veiled headdress (e.g., ibid., p. 101 Figure 4.5) as opposed to vertical locks or braids (e.g., ibid., p. 85 Figure 4.1; p. 93 Figure 4.3). Unlike the fully molded examples, the composite females never include molded jewelry and only rarely produce evidence for incised jewelry on the arms (two possible arm fragments may have incised arm bands: WT 5 and WT 147, ibid., p. 103).

This pattern might be specific to WT-13, however. In comparison, at Khirbat al-Mudaynah, the site produced a variety of hair styles on both its fully molded and pillar figurines, ranging from simple locks on the shoulders to elaborate hair styles, to headdresses or “shawls”, to simple veils (Daviau 2022, pp. 258–59). Mudaynah also differs from WT-13 in that only three figurine fragments—a mixture of frontally molded and pillar styles—came from the site’s cultic building (Temple 149), indicating that figurine-related rituals were far more common in domestic and neighborhood spaces (Daviau 2022, pp. 261–62). Nor were these fragments found complete, but were already broken, embedded in two different soil layers that surrounded the central podium in temple room 108 and were overlaid with fallen debris. Prior to destruction, the room was also possibly reused for the defense of the city (Steiner and Daviau 2024, pp. 140–47). In other words, the degree to which pillar and non-pillar female figurine molds differ at WT-13 might be an outlier, as is the number of figurines represented at the shrine, and this may reflect the unique context of that particular site.

Another reason for the continuation of face molds may be the efficacy attributed to the molding of eyes, noses, mouths and, where applicable, ears. The resilience of head molds may suggest that the very process of molding itself was considered essential to the ritual efficacy of the figures, such that the molded heads could not be easily dispensed with. I have argued elsewhere that the mask and protomes tradition in the Bronze and Iron Ages might either form a precedent or at least represent the same ritual focus on facial molds imbued with power (Darby 2014, pp. 344–47; see also Press 2012, pp. 171–72). For example, in the case of the Mesopotamian cult image, texts suggest that, to be efficacious, the design must have been recognizable and at least perceived as adhering to divine-sanctioned patterns (e.g., Berlejung 2021, pp. 7–8). In a ritual context where the visual display and design consistency of an image is tied to its perceived efficacy, molds ensure that an artisan line is producing an object that can actually function in the supernatural realm.

7.3. Regional Correlations Between Molds, Pillar Styles, and Gestures?

Most regions surveyed here include more than one type of head mold in circulation simultaneously, with greater variety between designs in certain regions (like Ammon, the Jordan Valley, Philistia, and northern Israel) and less variation between sites in others (like Judah). Even where molded faces adhere to a dominant style group, as in Judah, there is no evidence that a site used only a single mold or mold family. Instead, it is common to see multiple molds used to produce figurines in the same period.

For example, at the City of David in Jerusalem, where strong spatial and temporal context data are available, molded heads E1/10527 (Figure 2(16); Gilbert-Peretz 1996, Figure 11:16, Pl. 2:17), E1/10143 (Figure 2(3); ibid., Figure 11:3; Pl. 2:3), and E1/8456 (Figure 2(8); ibid., Figure 11:8 Pl. 2:8) all differ, but all were used in the same area of the site at the same time (Stratum 12) and, in the case of the first two heads, the same locus (L 1604). Nor is this unique in the eighth century. In Stratum 10, or the end of the seventh century through 586/587, circulating simultaneously was an unusual imported head G/4471 (Figure 2(15); ibid., Figure 11:15, Pl. 2:15–16), as well as G/5618 with clay added to the sides of the mold (Figure 2(17); ibid., Figure 11:17, Pl. 2:18), and the typical JPF head G/5723 (Figure 2(1); ibid., Figure 11:1; Pl. 2:1), all in Area G. Thus, at the site level, this pattern suggests that the specific molding for facial features that results in variations in eye shape and orientation, the number of rows of curls, the size of the face, and painted details may reflect the presence of multiple artisan cultures and production lines and the molds to which they had access rather than differing identities or ritual implementations.

In most locations, specific gestures do not clearly correlate with particular face molds or styles of pillar manufacture (with one exception, discussed below). For example, at WT-13, given the overall similarity of molded heads, the head likely associated with Daviau 2017, Figure 4.5:1, with hands joined on the chest, was similar to the molded head of ibid., Figure 4.5:4 (Daviau 2017, p.101) holding a drum perpendicular to the body. Given their hollow construction, the Ashkelon torso holding its breasts (Press 2012, p. 73, no. 63) had a molded head, as did the figurine holding a baby (ibid., p. 71, no. 59). Though Ashkelon produced a slightly broader range of hairstyles than WT-13, the relatively simple veil is the most dominant, suggesting that it is unlikely that these two torsos had completely different head styles. Conversely, in three coterminous mass burials at Lachish, different implementations of head molds (though all typically Judahite with rows of curls covering the ears) are attached to bodies with the same gesture, though with varying implementations of the added breasts and with pillars executed slightly differently in each case (cf. Kletter 1996, 74. Bc.4.A, Patai 1967, p. 65 Photo 1; Kletter 1996, 76.Bc.1.C; Tufnell 1953, Pl. 28:11; Kletter 1996, 77.Be.3-4.F; Tufnell 1953, Pl. 28:13; Kletter 1996, 78.Be.2.G; Tufnell 1953, Pl. 28:10; Kletter 1996, 75.Bc.2?; Tufnell 1953, Pl. 27:4, and Kletter 1996, 79B+.3.B, Tufnell 1953, Pl. 27:8).

In sum, there is ordinarily very little reason to argue that a given head mold style correlates consistently with particular gestures or pillar-base styles. Base technology seems to have little correlation with the identity of the image, and gestures seem to be more likely associated with function than with identity, though predominant functions could vary by region or site. For example, the breast-holding gestures on composite females in Jerusalem contrasts with the absence of the same gesture on composite figurines at WT-13 (though it is present on fully molded figurines at that site). The variation in hairstyle and adornment on molded heads may be the clearest opportunity to view regional and site-level design variation.

7.4. Judahite Composite Females and Intersite Homogeneity

To that end, the homogeneity of the Judahite assemblage demands further consideration. Artisans in Judah may have lacked a strong molded figurine tradition in the Iron IIA, with less evidence for molded attached female figures then in Israel, the Transjordan, and Philistia, and very few fully molded figurines in the Iron IIB-C. Instead of combining already-established head molds with the new standing pillar, Judahite artisans may have established a new mold tradition, one that duplicated important aspects of the image (forward staring eyes, smiling mouths, standing unaided) but largely eschewed hairstyles and molded decoration, like the polis, the veil, vertical side curls or twisted braids, diadems, hair descending below the neck, or exposed ears.

The short, simple wig and almost complete absence of ears in the Judahite tradition deserves further comment, since this region contrasts with almost every other composite head tradition. Coming originally from Egyptian art, unlike the tripartite wig, which often indicates a divine or elite status, the short, round, tiled wig primarily appears on male royalty in the Third Intermediate Period, high-status female figures during the 22nd dynasty, and non-royal men and women throughout the Late Period (Tassie 2009, esp. pp. 464–67). The short wig also became common in Iron II Phoenician and Syrian art on seals and ivories (Darby 2014, pp. 339–40, with literature).

What might this stylistic coherence suggest? One explanation could relate to the status of the Judahite composite female. The Judahite composites seem to forestall an easy association with a major deity, which might otherwise be suggested if the molded heads borrowed the typical Hathor headdresses of the Late Bronze plaque figurines known throughout the southern Levant (Cornelius 2004), or the tripartite headdress. When compared with surrounding regions, the headdresses remain unadorned and simple, suggesting that the heads are not associated with elite socio-economic groups, perhaps indicating that a subordinate status was assigned to the composite female in Judahite conceptions of the figure. While the same might be posited for the simple veil seen on WT 13 figurines or many of the Philistine composite heads from Ashkelon (Press 2012, pp. 166–67), veils and head-covering are associated with elevated social status. Head coverings are mentioned along with various items of adornment in Isaiah 3:16–24 as a marker of higher socio-economic class, similar to the way they were described throughout the ancient Near East (Létourneau et al. 2022; Fales 2021; Van der Toorn 1995). This begs the question of whether the veil, like other aspects of adornment, was intentionally eschewed by Judahite artisans when producing their versions of the composite female.

What possible explanation might account for the heads so rarely depicting the ears in Judahite examples? To speculate once again, Gabbay (2015) has remarked on the significance of whispering into the ears as an aspect of ritual activation in Mesopotamian practice, particularly salient in the rite of transition that transforms an artifact or raw material into a divine entity. Perhaps the Judahite artisans omitted ears to avoid the implication that they were representing divine entities or that they were used as the primary recipients of petitionary prayer. Surrounding regions seem far less consistent in their treatment of the ears, which are frequently depicted, perhaps intimating that these concerns were less salient.

Another possible explanation might be the function associated with JPF hand-on-breast figurines. In contrast to the lack of clear correspondence between head mold types and gestures in other regions, in the few instances of Judahite composite females found in Philistine sites, the Judahite-style head mold is only found on pillar bases with the breast-holding gesture (Press 2012, p. 168), and this is the dominant style in Judah, proper. The pattern may suggest that the gesture on composite Judahite figurines correlates with artisan choices in mold design. If it is correct to suggest that the hands-on-breast gesture was associated with protection from malevolent forces, and especially to help facilitate healing in domestic ritual (Darby 2014, pp. 337, 369–74, 388–93), then another reason not to depict the ears might be to protect the figurine from the danger associated with hearing ghosts, which can cause severe sickness or portend future harm (e.g., Scurlock 2006, pp. 7–8). In Mesopotamian tradition, as in the tenth tablet of the Šumma ālu series, the various sounds of the house itself or the evil entities in the house are treated as portended evil. In the case of Tablet X, the wind, here associated with demonic presence, is of particular concern. To quote Rendu Louisel’s (2016, p. 296) treatment of the topic:

“In exorcistic incantations, doors, windows, thresholds, and other kinds of openings constitute a dangerous and liminal space between the inside and the outside of the house, between the private and the public place. Prone to drafts, these particular spaces are also perfect for the manifestations of demons. In ancient Mesopotamia, demons are noisy and some sounds may be identified as their main manifestation. The terrible cry ikkillu is likely to be understood as the demon-Clamor (of mourning), being a sign of the evil it represents, that is, death.”

So, too, the nineteenth tablet of the series addresses the dangers associated with other entities, especially ghosts, in the home (ibid., pp. 297–98). Given the association with protecting domestic and neighborhood spaces, Judahite composite females may have shared the forward staring gaze of surrounding regions but lacked any susceptibility to the malevolent forces plaguing home life.

7.5. Intensity and Production Organization

When compared with Ammon, Moab, Philistia, or Israel, the intensity of figurine production in Judah stands in sharp contrast. This pattern certainly characterizes the composite style females (and even more so if we were to include pillar bodies with pinched heads), but the trend appears to hold for anthropomorphic figurines in general (on Israel, see Kletter and Saarelainen 2022, p. 157). Kletter and Saarelainen argue that the chronological range of occupation in the different regions may be a central factor. In the case of northern Israel, the majority of figurines from the north come from contexts between the tenth and eighth centuries BCE. They argue that figurine production becomes more intensive from the eighth through the seventh centuries, which are better represented at excavated sites in Judah (ibid). In comparison, the continuation of composite figurine forms in Phoenicia and Cyprus in the sixth and fifth centuries (after they have ceased in Judah) might suggest the same—where occupation is undisturbed, composite figurines and anthropomorphic figurines in general are better represented in the archaeological record.

Still the intensity of production in Judah is unlike most other southern Levantine locations inhabited in the Iron IIB-C, as is the Judahite stylistic adherence, particularly in the smaller range of gestures and short hairstyle that covers the ears. The dominance of the pillar figurine form in Judah also stands in contrast with the much wider variety of anthropomorphic figurine forms in surrounding regions or even in parts of the Negev (e.g., Malḥata, Kletter 2015). It is tempting to hypothesize that the coherence in Judah suggests, at least by the Iron IIB, that sites and perhaps artisan lines in the Judahite system were more closely integrated than those in surrounding polities. In contrast, Philistia was largely constituted by independent cities, as was Phoenicia and, in both of these cases, the figurine repertoire, particularly molds and hairstyles, shows considerable site-level and inter-site variety.

This leaves the question of whether composite figurine makers across sites in Judah were creating a more unified ritual landscape than what is visible in surrounding regions where greater variety exists. There remains little evidence for centralization of production in Judah. Petrographic results are currently mixed, with several studies suggesting that production was undertaken by different pottery lines at each site, while one study suggests that Jerusalem or Jerusalem area clays may have had some influence at Nasbeh in particular. Nevertheless, because Jerusalem may have adopted the composite form slightly later than sites in the Shephelah or Negev (Darby 2014, pp. 254–56),3 the city, though producing a large number of composite figurines, likely came into any artistic dominance only at the end of the eighth century and into the seventh century.

If true, the basic iconography of the Judahite composite female was already determined outside the boundaries of the capital city or temple elite, reflecting a decentralized iconographic tradition. Early distribution in the Shephelah and Negev may also tentatively suggest that the spread of the composite female technique in Judah reflects Phoenician and/or north Israelite influence, as it does in nearby sites from Philistia, like Ekron, and in central Jordan. At the same time, the specific adoption in Judah seems to demonstrate a far greater iconographic and technical homogeneity between Judahite sites and production centers, ranging from Mizpah in the north (McCormick 2023) all the way to ‘Aroer (Thareani 2011, pp. 188–89) and Tel Malḥata (Kletter 2015, p. 550) in the south, even without any evidence of centralized control.

7.6. Geography Revisited

To what extent do different molds indicate different cosmologies and geographies and at what scale? At the regional level, it seems the composite female face molds are the most likely indicator of local and/or regional conceptions of ritual landscape. If the figures are viewed as ritual implements rather than depictions of particular personalities, then the stylistic details of their manufacture do create boundary lines at the level of region or even site. Key differences between locations might focus on the status of the composite female, with the possibility that they were considered more subordinate in Judahite sites.

Differences in mold traditions between most sites in most regions seem to suggest that the way the composite female is imagined varied between sites of the same region or polity. There seems to be greater similarities across some regions, like Ammon, the Jordan Valley, northern Israel, and Phoenicia (specifically as applies to the heads without additional applied clay side locks), perhaps suggesting that artisans adopted a more flexible and permeable ritual landscape in these regions. These regions also preserve distinctive traits. For example, while applied clay side locks are known infrequently on composite heads at sites like Mudaynah and Jerusalem, they do not replicate Phoenician style, and this technique for composite females remains uncommon outside of Phoenicia.

How artists in a region imagine the function of the composite female may also vary. The females holding offerings or implements or with arms upraised may have been imagined as part of a deified retinue providing offerings and veneration. Drum-holding composite females may invoke a number of ritual purposes, ranging from scaring away demons (Shehata 2023, pp. 254–55), to ritual laments for previous or portended evil (Shehata 2023, pp. 258–59; Gabbay 2023, pp. 407–8), to laments sung to purify or protect a space or cultic object (Pongratz-Leisten 2022, p. 433). In this way, drum-bearing composite figurines might be imagined as part of a divine court or as apotropaic images to protect the spaces where they were used. Judahite composites may have been imagined as primarily focused on exorcism and protection over homes and neighborhoods rather than as agents to placate deities. While all of the composite females share the protective gaze and benevolent smile, perhaps composite females presenting the breasts, if associated with health and blessing, were intended to support the occupants of a house rather than provide service in a divine court.

The extent to which the composite tradition reflects the same ritual landscape or pantheon of a region’s inscriptional evidence might depend on whether or not the figure was understood as part of a retinue. In cases where the figurines holding implements or displaying gestures of supplication are found in actual shrines, like Sarepta or WT-13 (Daviau 2017, p. 275), clearly their presence is integrated in the shrine’s pantheon and with the deities of a region. One could argue, then, that the creation of a composite female in service to a local deity/ies would further reinforce those deities’ territorial boundaries, even if the same style of figurine was found outside of the shrine space. If a composite female was considered as an object ritually imbued with efficacy to exorcise and protect rather than serve a deity, then its integration in a territory’s belief system would depend on the ritual prayers and deities supplicated in accompanying ritual actions, as well as on perceptions about the malevolent forces the figurine holds at bay (cf. with Wiggermann 2000).

In sum, the free-standing and thus transportable nature of the bodies and the polyvalence and perhaps intended ambiguity of the composite design allowed various regions, sites, and even artisan lines to make use of the overall protective and mediating function associated with the shared features of the composite female while adapting the style for particular ritual needs best suited to location, circumstance, and belief system. The makers of composite females created an inter-regional Levantine consistency, as well as unique geographies and ritual landscapes, in reflecting diverse beliefs regarding the status of the composite female, her functions, and the local supernatural entities with whom she may have been integrated.