A New Analysis of YHWH’s asherah


Over the years the inscriptions found at Kuntillet ‘Ajrud (KA) and Khirbet el-Qom (KQom) that mention an asherah in association with the god YHWH have attracted a staggering amount of discussion and debate, and when considered as a whole what is most striking about this scholarly discourse is the degree to which it has been marked by persistent and wide divergence over the meaning of consonantal ʾšrth = אשרתה. While the prospect that the term may constitute evidence that a female deity was worshiped alongside YHWH in ancient Israel-Judah has been generally acknowledged, the question of whether it refers directly to a goddess has been hotly disputed, with not a few arguing that an object of some kind rather than a deity is in view.[i]   As a result, until now no one scholarly analysis has been able to catalyze anything approaching a consensus on the subject of YHWH’s asherah; instead interpretations have proliferated and alternative theories proposed, so that today several competing approaches to understanding ʾšrth exist, each one viewing the evidence significantly different from the others. Surveying this fragmented and disjointed scholarly landscape, it is difficult not to be reminded of the proverbial elephant in a dark room, where many individuals examining the same object identify it differently because of the particular kinds of data they take into account.

At this juncture it may be helpful to step back and freshly examine the issue without loyalty to any preexisting theory. Simply running in one of the tracks that has gone before is not likely to break the current interpretive impasse. What we need is a new approach that seeks to incorporate a greater range of data and evaluates prior theories both with regard to their legitimate insights and methodological deficiencies.

Furthermore, it would seem that the time is ripe for such an analysis. With the publication of the final excavation report of KA, we are now in a much better position to make informed judgments about the inscriptions found there and their relationship to the immediate epigraphic, iconographic, and material context. The readings at least of the inscriptions that mention YHWH and ʾšrth are for the most part well understood and epigraphically confirmed (Aḥituv, Eshel, and Meshel 2012: 73-142), while interpretations of the historical background and function of the site have become increasingly more judicious and balanced in their assessment (Zevit 2001: 370-381; Schmidt 2002: 96, 98-99; Frevel 2008: 39-40; Na’aman and Lissovsky 2008: 186–203; Mastin 2011: 69-85; Na’aman 2011: 300-302; Mandell 2012: 131-162; Meshel 2012: 65-69). At the same time, our ability to situate the inscriptions and other cultural phenomena from ancient Israel-Judah within the broader landscape of the southern Levant and eastern Mediterranean has developed considerably over the last several decades. Reconstructions of Israelite-Judahite religion have gradually become more rigorous in their application of historical methodology, which in practice has meant less bibliocentrism and more dependence on what can be clearly defined as primary sources of information, while various disciplines and fields of study on regions and cultures geographically and temporally proximate to Israel-Judah have made independent progress, allowing for more refined cross-cultural comparative analysis.

The following paper aims to take advantage of this state of affairs by reexamining inscriptional ʾšrth in order to critically assess previous scholarship on the issue and advance a new understanding of the term that builds on that same scholarship. In so doing, I hope to lay the groundwork for an interpretation that is not only methodologically defensible, i.e. less problematic than current alternatives, but historically more plausible given our present knowledge of the cultural and historical context of ancient Israel-Judah. I will begin by laying out my reading of the relevant inscriptions and then move directly to a consideration of the divine status of Hebrew ʾšrth as a necessary preliminary to the interpretive process of elucidating the meaning of the term. As we will see, if it can be established that ʾšrth refers to a divine entity, then this will have important implications for the possible range of semantic nuances that we may consider attributing to it. From here, we will then survey the three main interpretations of ʾšrth operative in current scholarship that assume the word/expression denotes a female deity. By carefully examining each of these proposals, we will be able to eliminate some readings of ʾšrth as historically doubtful and build on others as a means of proposing the hypothesis that the term asherah was a common noun used in monarchic era Israel-Judah to designate the female partner of YHWH.



The Inscriptions


brkt. ʾtkm.

lyhwh. šmrn. wlʾšrth.[ii]


“I bless you to YHWH of Samaria and to his asherah


brktk ly

hwh tmn

wlʾšrth. yb

rk. wyšmrk

wywy ʿm. ʾdn



“I bless you to YHWH of Teman and to his asherah. May he bless and protect you and may he continue with my lord”


]lyhwh htmn. wlʾšrth.  

]kl ʾšr. yšʾl. [mʾš?]. ḥnn hʾ wʾm pth wntn lh yhw klbbh.[iv]


“to YHWH of Teman and to his asherah[…] all that he asks (from YHWH?), he is generous, and if he entreats, then YHWH will give to him according to his wishes”[v]


[y]ʾrk. ymm. wyśbʿw[…] ytnw. l[y]hwh[.] tymn. wlʾšrth[

]. hyṭb. yhwh. hty[mn…]y. hyṭb. ym[m[vi]


“he will lengthen their days and they will be filled and they will give to YHWH of Teman and his asherah[…] do good, YHWH of Teman[ …] make their days good”


brkt ʾryhw lyhwh

wmṣryh lʾs̆rth hwšʿ lh






“Blessed be Uriyahu to YHWH and to his asherah, save him from his enemies…. to Abiyahu…. and to his asherah…. and to his asherah[viii]



The Divine Status of YHWH’s Asherah

When approaching the conundrum of ʾšrth, the first question we face is methodological: How do we go about interpreting a word in inscriptions whose meaning is uncertain and that appears in an expression for which no precise parallel exists elsewhere in Hebrew? Understandably enough, in the past the standard approach has been to accept a straightforward analysis of ʾšrth as the feminine noun ʾšrt with an attached 3rd m. sg. pronominal suffix –h and then to establish the term’s meaning from the vantage point of the known lexical-semantic values of asherah in other Northwest Semitic texts. Because asherah is clearly attested as the proper name of a major Levantine goddess, a heterodox cult object in the Bible, and even a shrine or cult place, each meaning is typically considered in turn and assessed as to which is most amenable to the inscriptional and religio-historical context with arguments advanced in favor of one or the other.

However, the problem with this philological either/or approach is that it treats the process of interpreting inscriptional ʾšrth in a rather mechanistic fashion, as though it were merely a matter of weighing and choosing among several well-defined possibilities. By focusing chiefly on asherah as proper name and asherah as cult object the analysis fails to recognize that the interpretive options are not only highly incommensurate to one another (divine/non-divine), stem from divergent ideological and cultural contexts (ancient inscription/traditional religious text), and may provide an incomplete picture of the semantic range of the term asherah as it was used during the Iron Age of ancient Israel-Judah, but the process of elimination that this approach entails invariably results in a distortion of the underlying evidence, so that some data is ignored or suppressed while other aspects are foregrounded and treated as decisive. For example, the interpretation of ʾšrt as a reference to the goddess Asherah is able to account for evidence that the inscription has in view a female deity paired with YHWH as an object of blessing but at the same time fails to adequately explain the significance of the attached pronominal suffix, while the interpretation of ʾšrt as a cult object belonging to YHWH resolves the pronominal suffix and yet downplays evidence that the blessing is directed toward a deity.

Methodologically, it would therefore seem wise to try to separate the issue of determining the divine status of ʾšrth from clarifying the precise meaning of the expression/term. Different and largely independent evidence bears on each question, requiring that the indications ʾšrth is a deity be given full consideration without feeling the need to prejudge the issue by fitting the term into a particular lexical-semantic slot. If it can be decided that ʾšrt refers to a deity on independent grounds, then this will provide a more reliable basis from which to theorize and speculate about its possible meaning.

With these prefatory remarks in mind, we can begin by pointing out that the argument in favor of interpreting ʾšrth as a deity is in fact functional in nature and has fairly little to do with the lexical-semantic value of the term asherah as used in other Northwest Semitic texts. The argument combines a number of factors both internal and external to the inscriptions, which can be listed in the order of their importance:


1) In the context of the inscriptions ʾšrth is invoked parallel with YHWH as an independent object of blessing. The parallelism is marked syntactically by the l- attached to both YHWH and ʾšrth and by the coordinating w- that separates them. Regardless of the presence of a possible suffix on ʾšrt, the syntax of the blessing implies that YHWH and ʾšrth are corresponding divine entities (Dever 1984: 30; Müller 1992: 28; Frevel 1995: 20-21; Zevit 2001: 404).


2) In comparable blessings from the broader region only deities are named as objects of the formula brk l-: e.g. brktk lyhwh “I bless you to YHWH” (Arad); whbrktk lqws “I bless you to Qws” (Ḥorvat Uza); brktk lbʿl ṣpn wlkl ʾl tḥpnḥs “I bless you to Baal Zaphon and to all the gods of Tachpanchas” (Saqqara); brktky lptḥ “I bless you to Ptah” (Hermopolis); brktk lyhh wlḥnb “I bless you to YHWH and to Khnum” (Elephantine) (Margalit 1990: 276; Müller 1992: 28; Pardee 1995: 302; Frevel 1995: 20-21; Tropper 2001: 101; Zevit 2001: 404; Rösel 2003: 107-121; Pardee 2005: 282; Leuenberger 2008: 121 n. 35).


3) As a number of scholars have argued, inscription 3.1 on pithos A is linked to an illustration of what appears to be YHWH and his consort (Gilula 1979: 129-37; Margalit 1990: 274-78; Coogan 1987: 119; Schmidt 1995: 96-102; Zevit 2001: 381-89; Schmidt 2002: 107-108; Thomas 2016).


4) The immediate archaeological context of the inscriptions at KA was evidently polytheistic. The divine name Baal is attested in at least two separate inscriptions, and El and “Name of El” are mentioned in a mythological context in a plaster wall inscription from the bench room complex (Dijkstra 2001: 24; Zevit 2001: 374, 404, 437; Aḥituv, Eshel, and Meshel 2012: 133).


5) There is growing evidence for the worship of female deities in Iron Age Israel-Judah, including widespread use of pillar figurines, cultic dualism in the form of standing stones, and other pictorial imagery, such as an incised image of a god and goddess pair recovered from eighth century Jerusalem (Kletter 1996; Dever 2005; Gilmour 2009).


6) The lexeme asherah is often associated with female divinity in ancient Syria-Palestine, including in the HB (Wyatt, 1999: 99-105; Day 1986: 385-408; 2000: 42-48).


In line with these various considerations, scholarly acceptance that asherah refers to a deity in the context of the inscriptions is much broader and more widely held than ideas about what ʾšrth means in particular or how the name should be analyzed. As we will see later, some scholars take ʾšrth to be the proper name Asherah with an attached suffix,[ix] others that the –h is integral to the spelling of the name,[x] and still others that the construction refers to a common noun asherah meaning “goddess” or “consort.”[xi]

The main alternative to interpreting ʾšrth as a deity is the theory that it refers to a cult object of some kind. This argument has been advanced in various forms since the beginning of research on the inscriptions and was recently featured as the preferred interpretation of the final publication of the inscriptions from KA (Emerton 1999; Day 2000; Smith 2002; Smith 2011; Aḥituv, Eshel, and Meshel 2012).[xii] The two points of data that have been critical to its development are the fact that inscriptional ʾšrth is most easily analyzed as ʾšrt + pronominal suffix h- and therefore seems to have in view a common noun rather than a proper name, and secondly, the equivalent of ʾšrt is commonly attested in the HB as a wooden cult object connected to illicit Israelite worship, namely the asherah (e.g. Ex 34:13; Deut 7:5; 12:3; 16:21). Depending on one’s interpretation of the biblical evidence and reconstruction of the historical context, proponents of the theory diverge into those who see the cult object as a representation of the goddess Asherah, either non-anthropomorphic or anthropomorphic (Gilula 1979: 134; Olyan 1988: 31-32; Ackerman 1992: 62-66; Emerton 1999: 335; Day 2000: 52-61; Sommer 2009: 45-46; Walls 2016: 275), and those who regard it as a non-anthropomorphic symbol connected to YHWH that need not have represented a female deity (Lemaire 1984: 42-51; Tigay 1986: 26-29; Keel and Uehlinger 1998: 228-232; Wiggins 1993: 181; Hadley 2000: 83; Smith 2002: 118-133; Mastin 2004: 326-351; Aḥituv, 2008: 221-24).

Yet despite the long history of this cult object reading, its methodology of interpreting the inscriptions on the basis of a particular Hebrew usage in the Bible is problematic. First, by focusing on the biblical meaning of asherah and how it resolves the difficulty of the attached suffix the analysis pays insufficient attention to what should be the primary context for determining the meaning of ʾšrth: the inscriptions themselves, and outside of that, the cultural and ideological world from which the inscriptions emerged. As was shown above, that context suggests rather unambiguously that ʾšrth is a deity. Several factors converge on this point, the most important being that asherah is treated on a par with YHWH and functions as a divine object of blessing.

Second, even if we were to accept the argument that the presence of the suffix militates against understanding ʾšrth as the proper name of the goddess Asherah, this grammatical feature cannot be used to exclude interpreting ʾšrt as a reference to a deity. Syntactically speaking, the pronominal suffix is strictly neutral on the question of the divine status of asherah. It tells us only that in mind of the speaker the entity asherah is sufficiently indeterminate, i.e. non-unique, to require it being specified for the linguistic-communicative context. To infer more than that is to go beyond the evidence.

Third, because the suffix does not by itself preclude interpreting ʾšrth as a deity, the main argument that the theory relies on to make the claim that the asherah is a non-divine entity is a lexical-semantic one: the asherah cult object meaning is attested in Hebrew and in a general way seems to fit the cultic setting (e.g. Tigay 1987: 174; Emerton 1999: 320; Smith 2002: 119-125). But this lexical-semantic argument is fundamentally inadequate for substantiating such a claim. Throughout modern research of the ancient Near East new or unknown deities are not primarily identified on this basis, i.e. whether their name conforms to some previously known divine nomenclature, but rather by their function and whether they are treated as divine by those who regarded them as such. In other words, we know a deity when we meet one because of the particular context in which it occurs, not because it bears a certain name. In addition, it is well known that the names and designations used for deities in many ancient Near Eastern cultures were highly variable in function and origin, embracing proper names, common noun titles, and nominal forms that otherwise denote non-divine objects or abstract concepts (cf. Rahmouni 2008). So it is questionable to assume that just because a nominal form refers to a non-divine object in one context that it must do the same in another.

Fourth, the preserved canon of Hebrew scripture is not a reliable lexical-semantic basis from which to make judgments about what meanings of the term asherah obtained in the context of ancient Israel-Judah (Coogan 1987: 119; Xella 2001: 72-74). While there is little reason to doubt that the biblical understanding of asherah as a cult object is somehow related to the asherah of the inscriptions, we cannot automatically assume their equivalence. The Bible is on the whole a much later document that took shape in an ideological and cultural context far removed from the world of the inscriptions, namely, a rigorous monotheizing and aniconic form of Judaism that came to prominence during the Second Temple period (Schmid 2003; Edelman 2009; Niehr 2010; Kratz 2015). Virtually all references to the asherah cult object occur in Dtr or Dtr-related literature, where it is portrayed as an illicit cult item and something that YHWH requires to be destroyed, whereas references to proper deities are few and far between.[xiii] It is thus unclear whether the biblical understanding of asherah as essentially a material cult object preceded the construction of this literature and had always belonged to the cultic vocabulary of Israel-Judah or whether it was an innovation peculiar to the Dtr authors, reflecting their agenda to rhetorically delegitimize cult statuary (Binger 1997: 138-41).

Furthermore, the cult object understanding of asherah is not the only meaning of the term preserved in the Bible. As we will see later, there are actually several different senses attested, including asherah as proper name of a goddess (e.g. 1 Kgs 15:13; 18:19; 2 Kgs 21:7), asherah as a class of female divinities (Jdgs 3:7), and asherah as illicit cult object (e.g. Ex 34:13; Dt 7:5; 12:3; 16:21). As a consequence, even if we were to agree that biblical Hebrew could serve as a guide for interpreting inscriptional asherah, one would still have to explain why the cult object understanding of asherah fits the context better than all other attested meanings.

Fourth, an auxiliary argument sometimes used to support the assumption that asherah is a non-divine entity is that all the continuations of the blessings and additional praise and prayer invoke only a singular masculine subject. For example, inscription 3.6 on Pithos B includes a prayer after the initial blessing in 3rd m. sg., “May he bless and protect you and may he continue with my lord.” The KQom inscription similarly ends with a prayer: “and from his enemies deliver (הושע) him!” with “deliver” in the 2nd masculine singular. From this a number of scholars have concluded that only YHWH is envisioned acting on behalf of the party who is blessed, rather than YHWH and asherah together. For example, Aḥituv, Eshel, and Meshel state, “The asherah is not incorporated into the blessings and prayers. The manner of use of the word אשרה in these formulas does not support the notion that it refers to a goddess” (2012: 132). However, because YHWH is a major focus of the discourse surrounding the blessings proper it does not follow that the asherah is non-divine or negate evidence that the term functions as a separate deity next to YHWH. If it were the case that the authors of the inscriptions thought only of YHWH as a source of blessing, then there would have been no need to mention the asherah with the w- and l- formulation.

In order to explain this aspect of the blessings, it is necessary to recall that ancient Israelite-Judahite culture was decidedly patriarchal and that in the public arena women were typically subordinate to men and particularly their husbands. Assuming a homology between the divine and human world, it would make sense for YHWH to be treated as the dominant authority and active partner in a male-female pair, that is, the one more than likely to be invoked and celebrated in public cult as the power that answered prayer and bestowed blessing on individuals in the community (Xella 2001: 73). Further illustration of the patriarchal conceptions underlying the focus on YHWH is found in several other aspects of the inscriptions in their religio-historical context: 1) The inscriptions all follow a formula of placing YHWH as the initial object of blessing, with the geographical specification attached to him if there is one, and then the asherah linked to him through a masculine pronominal suffix, the implication being that YHWH is primary and his asherah secondary. 2) As I have discussed elsewhere, the image of YHWH and his consort on Pithos A itself portrays YHWH in preeminent position, staggered in front of his female partner and significantly taller and provided with a larger crown. 3) Only male deities are mentioned in the other inscriptions recovered from KA, while epigraphic personal names found throughout late Iron Age Israel-Judah have tended as a rule to contain only male theophoric elements. Very few personal names have been discovered that invoke a female deity, even though we can be very confident that goddesses were recognized and venerated.[xiv]

So if patriarchy helps clarify the prominence of YHWH in the inscriptions, why the abrupt shift from a male-female dualism to a masculine singular subject? Relatively little discussion has been devoted to this problem by those who accept that asherah refers to a deity (cf. Dever 1984: 29-31; Coogan 1987: 118-119; Uehlinger 1997: 140-142; Zevit 2001: 397; Naʾaman 2011: 303-304). H. P. Müller suggested that the blessing for general wellbeing directed to YHWH and his asherah in inscription 3.6 should be functionally distinguished from the succeeding prayer for guidance and protection, which he sees as relevant only to YHWH (1992: 32). But it is difficult to see how the content of the invocations is all that different (brkty/ybrk), and in any case this explanation fails to reckon with the broader pattern of invocations of YHWH and his asherah followed by verbal content in the masculine singular, as we saw above. Another interpretation suggested by various scholars is that the masculine singular language has in view YHWH and his asherah acting together as a compound entity, with the agency of the asherah having been subsumed into the male god’s agency (Frevel 1995: 20-21; Schmid 2003: 24-25). In her treatment of the inscription at KQom, J. Hadley noted, “As Yahweh (and) his asherah can be regarded as a compound subject, a plural verb is not necessary (cf. especially Kuntillet Ajrud). There are several instances in the Old Testament where a compound subject takes a singular verb (cf. e.g. Prov. xxvii 9 and xxix 15)” (2000: 96). Although Hadley herself did not take asherah to be a divinity, her comment would still nevertheless apply in this case.

Furthermore, as first recognized by C. Frevel (1995: 21), we have a fairly close parallel to the phraseology invoking YHWH and his asherah and then YHWH alone in Punic and Late Punic dedicatory inscriptions, which sometimes begin by invoking Baal Hammon and Tinnit together and then conclude with a formula in the masculine singular, “To the Lord Baal Hammon and to our Lady Tinnit the Face of Baal… because he heard his voice, he blessed him” (KAI 102).[xv] With regard to the ending formula, Emerton has expressed doubt that the reading “he blessed him” is correct, since the orthography of 3rd m. sg. forms of the verb brk “to bless” is often indistinguishable from the plural (1999: 320-21). But according to Robert Kerr, the singular number of the verbs in the formula is shown by the Punic dedication in Greek script from El-Hofra (KAI 175; σαμω κουλω βαραχω) “that can only render ‘he heard his voice, he blessed him’ [and] is supported by the same formula translated into Greek in KAI 176; this is also supported by some vocalized neo-Punic spellings” (personal communication, Dec. 28, 2013). Thus in the space of a single inscription the divine object of dedication shifts from speaking about two separate male and female deities to an emphasis on the masculine principle only, as though the latter were corresponding to or inclusive of the former. Interestingly, Kerr’s explanation for the shift to the masculine singular is that Tinnit “when mentioned in such texts is invariably denoted as the /fanḗ bâl/ or literally ‘the face of Bal’ and not as a [completely] independent entity. Tinnit is thus an aspect of “Bal (H)Ammon.”

Based on this parallel for a divine male-female pair being treated as a functional unit, it is fairly easy to see how “YHWH and his asherah” could be understood similarly. On the assumption that ʾšrth consists of ʾšrt + –h masculine suffix, the asherah is linked directly to her male partner as an entity subordinate to him, comparable to the linkage of Tinnit to Baal through the epithet “Face of Baal.” By associating the asherah to YHWH in this manner, not only is YHWH’s agency and authority over the asherah emphasized, but the two are portrayed as symbiotically and closely interrelated, making the independent female agency susceptible to being expressed as integral to YHWH. If this interpretation is correct, then it would mean that when an individual blessed another person to “YHWH and his asherah” and then continued with prayers in the masculine singular it was not a case that the agency of the female deity was seen as non-existent or ineffectual, but rather that she was implicitly assumed to have a contributory role in the matter and regarded as somehow inseparable from YHWH (cf. Levine 2014: 182).

Finally, the cult object theory assumes that a blessing by a deity’s cult object, as distinct from the deity, was conceivable in an ancient Near Eastern context. That is to say, according to this view, asherah was understood to be essentially the name of a cult emblem, whether it belonged to Asherah or YHWH. But as has been pointed out by others, in the ancient world the deity and cult image were substantially coterminous. The image was deity and deity the image (Olyan 1988: 31; Ackerman, 1992: 65; Xella 2001: 71-81; Schmid 2003: 23-24; Parker 2006: 87-88; Merlo, 2009: 979; Levine 2014: 182-185). Although proponents of the theory have searched far and wide, they have not been able to produce one convincing example of a blessing by a cult symbol that was not in itself already considered divine.[xvi]

For example, J. Tigay (1986: 28-29) has cited several supposed parallels to the notion of blessing by a cultic object or sanctuary, including the Babylonian epistolary salutation, “may Uruk and Eanna bless my lord”; a Phoenician inscription from Byblos recording a dedication to “our Lord and the Image [sml] of Baal” (KAI 12); and memorial graffiti from Hatra directed to various deities as well as smytʾ “images” (KAI 251; 256). But closer examination of these expressions casts severe doubt on his analysis (Xella 2001: 74). The first blessing by Uruk and Eanna has a peculiar history tied to southern Babylon, where temples and cities were sometimes treated as divine subjects or actants. Bill Arnold has shown how the epistolary salutation was a “gubernatorial formula employed only by the governors of Uruk during the seventh century B. C.” (1992: 384), whereas Parpola’s compilation of letters from Assyrian scholars to Esarhaddon and Assurbanipal reveals that the norm in northern Mesopotamia was to direct blessings to deities, generally Nabu and Marduk (1983: 439-42). With regard to the dedication to the “Image of Baal,” this is likely an epithet of a female deity who was companion to the male deity referred to as “our Lord.” J. Teixidor has written that the epithet “certainly refers to the Lady of Byblos, and is similar to the epithets ‘Face of Baal’ and ‘Name of Baal’ borne by Astarte elsewhere in Phoenicia” (1977: 48). Lastly, it appears that the “images/standards” invoked at Hatra were divine as well, based on the memorial context and coordinating w- linking them to the preceding list of deities, “our Lord, and our Lady, and the Son of our Lords, Allat …” (Gudme 2013: 111).

The other parallels sometimes brought up in discussions of asherah as a cult object are cases of individuals swearing by a temple or cult paraphernalia in a fifth century text from Elephantine (AP no. 44/TAD B 7.3) and in biblical, early Christian, and rabbinic sources (cf. Lemaire 1977: 608; Tigay 1990: 218; Emerton 1999: 333; Aḥituv, Eshel, and Meshel 2012: 132; Aḥituv 2014: 35-36). However, the full context of the oath from Elephantine reads “Oath of Menahem… that he swore to Meshullam… by [YHW the god] in the sanctuary and by Anatyahu” (mw[mʾh zy] mnḥm… ymʾ lms̆lm… by[hw ʾlh]ʾ bmsgdʾ wbʿntyhw).[xvii] Αlthough the preposition b- attached to msgdʾ could be interpreted grammatically as “in” or “by,” several factors show that in this context its significance is locational, including 1) the absence of a coordinating w- before bmsgdʾ, which suggests that the msgdʾ performs a function in the oath different from ʿntyhw; 2) extensive comparative evidence that oaths were customarily rendered in cultic settings (van der Toorn 1995: 2051; Chaniotis, 2009: 128; Sandowicz 2012: 93-94; Sommerstein and Torrance 2014: 132-138; Hurowitz 2015: 389-418); and 3) comparison with AP no. 6/TAD B 2.2 that speaks of an oath made byhw ʾlhʾ byb, “by YHW the god in Yeb” (line 4). The inclusion of this locational information thus served to specify that the oath was made to a particular manifestation of YHW, highlighting the local and communal character of oath taking.

On the other hand, credible references to a practice of swearing/blessing by a temple and/or non-divine cult paraphernalia appear only in very late sources. Against Aḥituv, the context of Amos 8:14 suggests that the ʾas̆mat of Samaria is best understood as a reference to a local manifestation of YHWH (Olyan 1991: 121-49; Paul 1991: 270; Toorn 1996: 322; Cogan 1999: 105-106), which means that it provides further witness to the convention of directing oaths to deities. This leaves only Matt 23:16-22 and miscellaneous rabbinic traditions about swearing by the temple and altar, which are so far removed from eighth century KA and reflect the peculiar theological interests and zeitgeist of the discourses in which they appear that they can hardly be considered relevant (Zevit 2001: 404, n. 114). Because in these cases the temple or altar function as a stand-in for the holiness of the one God, it seems reasonable to suppose that the practices underlying the textual descriptions originated as a development or aniconic variation of earlier customs of swearing by a deity’s cultic manifestation, as seen in the oath from Elephantine.

In sum, that inscriptional ʾšrth refers to a female partner of YHWH is the most plausible interpretation of its function in the context of the blessings. The fact that it is treated by the authors/speakers as though it were a deity must be given priority to other grammatical and lexical-semantic considerations, such as how ʾšrth should be analyzed or what specific nuance of meaning it may have.


The Meaning of ʾšrth

When it comes to the question of what ʾšrth exactly means and the identity of the deity referred to by it, the evidence is more ambiguous and uncertain. The proposals that have been offered to date fall into three categories: 1) ʾšrth is the name of the goddess Asherah with an attached pronominal suffix; 2) ʾšrth refers to the goddess Asherah, but with a different linguistic form; the he is integral to the spelling of the name; and 3) ʾšrth reflects asherah with an attached suffix and refers to a deity distinct from Asherah. The term has the semantic valence of a common noun, such as “goddess” or “consort.”




How to decide among these alternatives? The first has been the most popular among scholars who accept that ʾšrth refers to a goddess. Their reasoning is as straightforward as it is transparent: a) Asherah is well-attested as a divine name in biblical and extra-biblical sources; b) archaeological evidence for goddess worship in ancient Israel-Judah is sufficiently clear; therefore, c) inscriptional ʾšrth should be translated as “his Asherah” (e.g. Dever, 1984: 21-37; Coogan 1987: 115-124). However, the major difficulty with this line of interpretation is that it fails to explain the presence of the suffix and its implication for understanding the meaning of asherah. What would it mean to say that the goddess Asherah is YHWH’s Asherah? As has been repeatedly observed, this approach would seem to make vacuous any notion of proper names.

One possible way of getting around this grammatical difficulty is to beg the question of whether divine names functioned as regular proper names in Israel-Judah and to argue that while the attachment of a suffix to a divine name is unprecedented in terms of known Hebrew usage it may nevertheless have been possible in the cultural world of the inscriptions (Freedman 1987: 241-49; Müller 1992: 15-51; Uehlinger 1997: 140-142). The rationale underlying the standard objection to interpreting inscriptional asherah as the name of a goddess is that divine names had the grammatical determinacy of proper nouns and as such entailed the same functions and uses. For example, J. A. Emerton has argued that “the use of a suffix with a personal name is not in accordance with Hebrew idiom as far as we know it, and it is unwise to interpret the newly-found inscriptions in such a way unless there is no satisfactory alternative” (1982: 14-15), implying that the divine name Asherah is a personal name, just as David, Abraham, or Ruth. But is this assumption correct? Do we know that divine names were proper names in the sense that they were used to designate conceptually unique entities? Were deities comparable to individual human beings in terms of their need for grammatical specification?

Coming from our modern Western context, it is difficult to imagine a religious worldview in which deities are not conceptualized as unique supernatural characters whose identities could be called to mind through simple first names. We naturally expect deities to be comprehended as defined anthropomorphic personalities that transcend the mundane world, and so tend to be drawn to such representations when we encounter them in the literature of ancient cultures from the Near East. As a consequence, the mythological, mytho-historical, and devotional texts of these cultures have until recently played a disproportionate role in influencing how modern scholars understand the way ancient Near Eastern peoples themselves conceptualized divinity. Because these texts portray deities as cosmic individuals who can be singularly located in space and time and whose names are used as though they refer to coherent unitary identities, we assume that they were always imagined as such, that their supernatural anthropomorphism was a defining aspect of their identity.

However, when the total range of ancient textual material is taken into account, we see that the tendency to think of deities in abstract and supernatural terms only tells a part of the story of how ancient Near Eastern peoples conceptualized divinity. Over the last half-century archaeologists have unearthed texts dealing with cult and religious worship from throughout the region and diverse eras and as a result it has become increasingly clear that in regular practice divine personhood was treated as materially-expressed, metaphysically complex, and multiform (Hundley 2013: 363-371; Porter 2009: 153-94). As B. Sommer has noted, ancient Near Eastern deities were fundamentally unlike human beings in that they were conceptualized as having multiple bodies that were simultaneously coexistent. Because deities could inhere in material icons of various sorts, their identity was fluid and fragmented (2009: 12-37). Their presence could be extended and spread over multiple representations through a statue-deity symbiosis, causing the divine person to be differentiated and pluriform rather than strictly unitary and discrete. Individual gods and goddesses could be replicated (theoretically endlessly, if the necessary rituals were performed), and each physical manifestation could legitimately be understood as belonging to the category elohim. Paradoxical as it may seem, a simple first name was not always sufficient to identify a particular deity.

Perhaps the most dramatic evidence of this manner of conceptualizing divinity in an Israelite-Judahite context is found in the inscriptions from KA where YHWH is named both “YHWH of Samaria” and “YHWH of Teman.” Although there has been some equivocation among scholars about what these designations mean (cf. Emerton 1982; Allen 2011), I think that there can be little doubt that they represent two distinct deities tied to two separate localities: a form of YHWH worshipped in the region of Samaria and another worshipped in the area of Teman-Edom (Smith 2012: 205-50; Hutton, 2010: 177-210; Schmid 2003: 25-28).[xviii] Other inscriptional and biblical evidence similarly shows that YHWH had multiple local manifestations. An inscription from Khirbet Bet Lei (Naveh 2001: 197-98) refers to YHWH as “the god of Jerusalem” and in the HB we have mention of “YHWH in Hebron” (2 Sam 15:7) and “YHWH in Zion” (Ps 99:2).[xix] Amos 8:14 speaks of people swearing by the Ashima/Name of Samaria, the god of Dan, and the Strength of Beersheba, which are probably best interpreted as epithets of local iterations of the god YHWH.[xx] Furthermore, all the above local forms contrast with the cosmic form of the deity attested in biblical texts and inscriptions as “YHWH of the armies,” an epithet that links YHWH to the sky as supreme commander of the heavenly hosts (Choi 2004: 17-28).

The implications of this data is that like their Near Eastern counterparts Israelite-Judahite deities had objective realities distinct from and in addition to their more individualized mythological personalities. Defying modern grammatical categories, deities could straddle the realms of proper and common nouns. For abstract mythological purposes, one could, and, indeed, had to speak of YHWH as a singular cosmic entity that transcended the mundane world and manifested his power over it: “who is like you among the gods, O YHWH” (Ex 15:11). For purposes of cult, however, where the material and tactile facilitated worship, it was necessary to distinguish which particular form of YHWH one was dealing with, resulting in the specifying epithets seen above.

Thus from a strictly materialist perspective we would have to conclude that the grammatical argument that ʾšrth cannot refer to the goddess Asherah because of the attached pronominal suffix lacks cogency and is based on questionable assumptions. When we set aside the specific question of whether divine names could carry pronominal suffixes and limit ourselves to an analysis of how deities were conceptualized and their personal names functioned in religious worship, then it becomes clear that Israelite-Judahite deities were not exclusively understood in the category of proper nouns, but were sometimes treated as non-unique entities whose identification required further specification beyond their simple first names.[xxi] To be sure, understanding divinity as pluriform and quasi-indeterminate does not necessitate the assumption that inscriptional asherah is the name of a goddess. We have not demonstrated that pronominal suffixes were a viable means of determining divine names or that “his Asherah” is the most plausible reading of consonantal šrth in the context of the inscriptions. But the evidence described above would at least allow one to make a theoretical argument in support of such an interpretation. Because YHWH could be treated as almost indeterminate, something in need of further definition in the context of cult, then it would logically follow that the same would have applied to his consort. If this is the case, then it is possible that a possessive suffix attached to asherah is performing a specifying function comparable to the geographical names attached to YHWH. Just as the latter are used to specify exactly which form of YHWH is the object of blessing, the possessive suffix may have been intended to specify a particular cult manifestation of the goddess via her association to a geographically specified YHWH. We will return to the question of whether this interpretation of ʾšrth is satisfactory shortly.

In addition to the evidence that Israelite-Judahite deities were pluriform and indeterminate, the other major line of evidence that has been advanced in support of interpreting inscriptional asherah as the name of a goddess is the use of divine names elsewhere in the ancient Near East in bound constructions syntactically parallel to YHWH and his asherah. P. Xella has collected and discussed examples from Ebla and Ugarit where female deities are linked to male deities through a pronominal suffix as background for interpreting the inscriptions from KA and KQom (1995: 599–610; 2001: 74-75). The examples cited from Ebla include “Kura and his Barama,” “Rashap of Adani and his Adamma,” “Rashap of Duneb … and his Adamma,” the latter two featuring geographical specifications parallel to YHWH of Samaria/Teman, whereas at Ugarit Anat is linked to the deity Gatharu in the forms “Gatharu … and his Anat” (KTU 1.43.13) and possibly “Anat of Gatharu” (KTU 1.108.6).[xxii] Significantly, Xella notes that at both Ebla and Ugarit the invocations appear in ritual contexts and most likely refer to the cult statues of these deities (Xella 1995: 610; 2001: 74-75).

Closer in time and space to ancient Israel-Judah are the bound forms Anat-Bethel in Neo-Assyrian texts listing oath deities (Parpola and Watanabe 1988: 24-27, 28-58), Ashtar-Kemosh in the Mesha stela (KAI 181), and Anat-Yahu, Anat-Bethel, Ashim-Bethel, and Herem-Bethel from the archives of Elephantine. Taken together with the evidence from KA, these constructions are indicative of a broader West Semitic cultic convention of divine possession, where one deity, generally female or otherwise subordinate, was seen as belonging or inseparably linked to another deity (cf. Smith 2001: 72-74; Zevit 2001: 403).

The divine names Anat-Yahu and Anat-Bethel from Elephantine may have particular relevance to inscriptional ʾšrth and deserve closer examination, as they are attested in the context of a Judean diaspora community and likely reflect a continuation of goddess worship from Israel-Judah. Anat-Yahu is mentioned once as the divine object of an oath (AP no. 44/TAD B 7.3) and Anat-Bethel, presumably a variant name of the same goddess, is described as a recipient of temple-fund shares for the temple of Yahu in Elephantine along with Yahu and Ashim-Bethel (AP no. 22/TAD C 3.15). There has been much debate over the meaning and origin of Anat-Yahu/Bethel (AP: xviii-xix; Porten 1968: 173-179; van der Toorn 1992: 80-101; Niehr 2014: 153). Some have questioned whether the figure represents an independent goddess and is not rather a cultic symbol or hypostasis of YHW, because of the predominance of YHW as the primary deity invoked in documents and personal names from the Judean colony (McCarter 1987: 147; Grabbe 2013: 127-28). But we have already seen that references to male deities tend to far outnumber those of female deities in the epigraphic record of Israel-Judah, so the disparity between invocations of YHW and Anat-Yahu/Bethel in the preserved texts from Elephantine is not really all that surprising. Furthermore, the divine status of Anat-Yahu/Bethel is indicated by several interrelated pieces of evidence, including 1) the tradition-history of the divine name Anat-Bethel and the interchange of Yahu in the position of Bethel in the compound name Anat-Yahu (van der Toorn 1992: 81); 2) the invocation of Anat-Yahu as a distinct deity in the oath mentioned above, where she is separated from Yahu by the w- and b- formulation; 3) the inclusion of Anat-Bethel with Yahu and Ashim-Bethel in the list of temple-fund donations; if Yahu is functionally correspondent to Bethel, then the structure of the names Anat-Bethel (Anat of Bethel) and Ashim-Bethel (Ashim of Bethel) imply that together they function as a divine triad, the quantity of silver dedicated to Anat-Bethel reflecting her subordinate status with respect to Yahu and seniority over the junior figure of Ashim-Bethel: Yahu receives 126 shekels, whereas Anat-Bethel receives 120 Shekels and Ashim-Bethel 70 Shekels (Röllig 1999: 174; Berlejung 2012: 205-207); 4) the use of Anat as an epithet for the consort of Bethel among the Aramaean community of upper Egypt, as shown by Papyrus Amherst 63 (Steiner 2003: 314 col. VII.7-19); and 5) the detail reported in the letter requesting permission to rebuild the temple of Yahu that when it was destroyed multiple pillars or sacred standing stones were demolished, implying the existence of a small pantheon of deities (Becking 2011: 140-41). All of these factors indicate that in the context of Elephantine Anat-Yahu/Bethel could have only been understood as a discrete female figure paired with YHWH.

With regard to the origin of Anat-Yahu, scholars are divided over when and how this concept developed. Some have proposed that the worship of this goddess by Judeans at Elephantine represents a continuation of monarchic Israelite-Judahite devotion to Anat (Day 2002: 143-144), whereas others have argued that the pairing of Anat with YHWH may have arisen through syncretism with local Aramaean culture in Egypt (Porten 1968: 173-79; Oldenburg 1969: 85-86), or that the Judeans of Elephantine had a cultural background in the Northern Kingdom where a syncretistic blending of Aramaean and Yahwistic cult had occurred while still in the land of Israel (van der Toorn 1992: 80-101; Niehr 2014: 153).

However, none of these hypotheses are very convincing. First, the notion that Anat-Yahu represents a continuation of earlier Iron-Age religion from Israel-Judah is undermined by the fact that there is little to no evidence that Anat ever functioned as a consort of YHWH. While the divine name Anat appears in a few personal names and toponyms in the HB, this usage is sporadic and of uncertain significance, perhaps hearkening back to an earlier second millennium cultural strata (Römer 2015: 86). Attempts to find evidence for a cult of Anat by emending various biblical texts or to see her standing behind the Queen of Heaven in Jer. 7 and 44 have not generally been accepted (Day 1999: 36-43; Smith 2002: 103; van der Toorn 1992: 81-83; Becking 2003: 221-224). On the whole, we lack positive indication that Anat was worshipped as a major goddess in Iron II Israel-Judah, whereas all available data from the Bible and inscriptions is that Asherah was chief goddess and that YHWH was paired with a goddess designated “his asherah.” Furthermore, this explanation of the origin of Anat-Yahu/Bethel fails to reckon with the strong Aramaean character of the divine name and the associated figure of Ashim-Bethel. As others have noted, Anat-Yahu appears to be patterned off Anat-Bethel, and the whole constellation of Bethel, Anat-Bethel, and Ashim-Bethel seems to have roots in North Syria far from the native homeland of the Judean colony (van der Toorn 1992: 83-85). The worship of Anat-Yahu at Elephantine therefore is unlikely to be a direct carry-over from earlier Israelite-Judahite tradition.

On the other hand, the theory of local syncretism is highly inadequate as a conceptual category for describing the veneration of Anat-Yahu at Elephantine, since it presupposes the biblical perspective that there was some earlier more pristine version of Israelite-Judahite religion from which the Judeans at Elephantine had departed. As van der Toorn has observed, the documents from Elephantine reveal a “Jewish minority group that is otherwise keen to preserve its native religious culture” (1992: 83), so it hardly seems credible that they would have adopted an entirely new goddess as a result of local influence. R. Kratz has identified the religious culture at Elephantine as a “standard manifestation not only in the Israelite-Samarian region but also in Judah itself,” whereas the biblical Judaism that developed in the Persian and Hellenistic periods in Jerusalem was the exception to the rule (2015: 143). J. Anderson has similarly written, “Since there is no evidence that the Jewish community at Elephantine knew the Torah, and through it the notion that the worship of Yahweh excludes other gods, it is more legitimate to view the Yahwism of Elephantine as traditional Israelite religion that predated the instauration of strict monotheism” (2015: 33).

The last theory attempts to mediate between these two positions by recognizing on the one hand that Elephantine goddess worship more than likely should be traced back to monarchic Israel-Judah and on the other that Anat-Bethel was originally a foreign Aramaean deity. Yet pushing the syncretism with Aramaean religion back to a period when the ancestors of the Elephantine Judeans were still living in Samaria is speculative and problematic as well. First of all, the cultural and religious background of the colony is clearly Judahite rather than Aramaean/Samarian. They refer to themselves as Judaeans, bear Yahwistic names, maintain contact with Judaean religious authorities, and hold to distinctive Jewish practices, such as Passover and Shabbat (Becking 2011: 128-42; Porten 1968: 105-150, 173-176). Recent research suggests that the origin of the colony should be traced back to the final decades of the Judahite monarchy, corresponding to the resurgence of Egypt as a major power in the eastern Mediterranean and when Judah was under its influence and there was conflict between the 26th Dynasty and the kingdom of Kush (Kahn 2007: 507-516; Botta 2009: 12-16; Redford 1992: 441-445, 462-463. Cf. Porten, 2003: 451-70; Becking 2011: 130-31).[xxiii] Nothing about the Elephantine Judeans indicates they were ethnically heterogeneous or that they had arrived far south on Egypt’s border as refugees from the Babylonian destruction of Jerusalem. They were simply a mercenary colony originally from Judah who spoke Aramaic and operated in an Aramaic-Egyptian multicultural environment and could be designated Aramaean for Persian administrative purposes (Botta 2009: 53-54).

Furthermore, the thesis that the group had a northern background from the region of Samaria is primarily dependent on biblical texts alleging that Israelites in Samaria had mixed their cultic traditions with those of foreigners imported by the Assyrians, adopting Aramaean deities such as Ashima and Bethel into their distinct form of Yahwism (e.g. 2 Kgs 17; Jer 48:13). However, closer analysis suggests that these texts are polemical in nature and of doubtful historical value. Archaeological investigation reveals that there was strong continuity in the material culture of Samaria and the surrounding highlands during the eighth-seventh centuries and that the number of foreign transplants was not very high: “Whatever exiles from foreign states were forcibly imported into the Samaria highlands, most seem to have been absorbed into the local population” (Knoppers 2004: 171). In addition, the evidence of proper names found on bulla, coins, and papyri from the Persian period suggest that the mass of people living in the province were Yahwistic/Israelite (Knoppers 2006: 275-278). No personal names with the Aramaean theophorics Bethel or Ashima are known (Cross 2006: 75–90; Zsengeller 1996: 182-189). The Deuteronomistic authors thus seem to be engaging in a kind of ideological warfare and identity building project of portraying Israelites/Samarians in the north as fundamentally impure and compromised by foreign influences.

Because of the Aramaean-like character of the divine names used by Judeans at Elephantine, the most likely explanation for Anat-Yahu, Bethel, and Ashim-Bethel is that the names were adopted by the colony in the course of its acculturation to the Aramaean milieu of Upper Egypt but at the same time this adoption was not viewed as a departure from earlier Judahite tradition or that the identities of the deities standing behind the names had altered in any significant sense. The phenomenon of cultures translating their deities to new cultural contexts is well known. At sites of strong intercultural contact divine names were often interchanged or updated in order to create links between different cultures (Smith 2010). Because goddess worship was native to Jewish Elephantine’s parent culture, it seems more sensible to assume that Anat-Yahu as a cultically worshiped figure ultimately derives from Judahite tradition, while her name is simply a locally acceptable way of referring to the consort of the pantheon head that was probably known by another name in Judah.

In light of the above, the fact that the Judeans at Elephantine refer to their main goddess with a bound construction seems relevant to understanding inscriptional YHWH and his asherah. As with other aspects of their cult, this practice may reflect continuity with Iron Age Israelite-Judahite religion and provide support for interpreting ʾšrt as the name of a goddess. Admittedly, too much weight cannot be placed on this evidence, since as we mentioned above Anat-Bethel is a name with a prior Aramaean history and Anat-Yahu was probably constructed on analogy with Anat-Bethel. However, if Elephantine Judeans were the ones responsible for the interchange of Yahu with Bethel in Anat-Yahu, as I think must have been the case, then it suggests that these sorts of bound constructions were at home in their culture of origin and not an innovation peculiar to Elephantine.

Further indication that the polytheism of Elephantine was closely related to Israelite-Judahite tradition is reflected in the text that mentions Anat-Yahu (AP no. 44/TAD B 7.3), which I have reconstructed to read “by [Yahu] the [god] in the sanctuary and by Anat Yahu.” The text is unfortunately worn away near the beginning of the line and the second letter is only partially preserved. Cowley read a yod and restored YHW, whereas Porten believed it was a ḥet and restored Herem (AP: 148; Porten 1968: 317).[xxiv] However, orthographically the letter shares more similarities with the initial yod seen in lines 2 (ימא) and 8 (יהב), and while the vertical tail is somewhat anomalous, it seems too short and thin to be a ḥet based on other examples in the text (lines 1, 7-9). In addition, contextual considerations lend support to reading Yahu over Herem. First, the only other instance in which Herem is invoked outside personal names is found in compound form, “Herem-Bethel the god” (AP no. 7/TAD B 7.2), which indicates that like Anat-Yahu, Anat-Bethel, and Ashim-Bethel this was the full cultic name of the deity. As Anat-Yahu appears in compound form in the oath, so we would expect the same of Herem-Bethel. Yet according to Porten’s reconstruction the gap between the first letter of the name and the final aleph of אלהא would have been insufficient to allow for the full name. Second, the only deity in the Elephantine documents attested with a geographical specification is YHW: AP no. 6/TAD B 2.2 speaks of an oath made byhw ʾlhʾ byb, “by YHW the god in Yeb” and TAD B 3.12 invokes yhw ʾlhʾ škn yb brtʾ “YHW the god who dwells in Yeb the fortress.” AP no. 44/TAD B 7.3 thus seems to parallel these other instances by localizing YHW to his sanctuary in Elephantine.

If this reading is correct, then the syntactical arrangement of the invocation of deities in the oath is redolent of the manner of invoking YHWH and his asherah at KA. As at ‘Ajrud, the primary deity in view is mentioned first: YHW/YHWH. After this is attached localizing information that specifies the manifestation of the deity: “in the sanctuary”/Teman and Samaria. In final position is the goddess, who is linked to her husband through a bound construction: Anat-YHW/his asherah.

So far our investigation has focused on two lines of evidence that would possibly allow for interpreting YHWH’s asherah in the inscriptions from KA and KQom as the personal name of the goddess Asherah. On the one hand, deities in the world of ancient Israel did not have the same determinate status as human beings and in certain settings required extra specification to be identified. On this view, the suffix attached to asherah could be interpreted as a means of specifying the precise cultic manifestation of the goddess. On the other hand, female deities were sometimes linked to male deities through bound constructions marking possession. These constructions include either a construct chain where the female is directly related to the male (female DN-of-male DN) or an extended conjunctive form that marks possession through an attached suffix (male DN-and-female DN + 3rd m. s. suffix).

We have already mentioned that a number of scholars have found one or both of these lines of evidence sufficient grounds for concluding that YHWH’s asherah is no less than the goddess Asherah. And indeed, if we assumed that Anat in the compound name Anat-Yahu were the proper name of a goddess, then the links between Elephantine and Israelite-Judahite religion described above would appear to make that conclusion even more attractive. From the perspective of linguistic function, the constructions “Anat of YHW” and “YHWH and his asherah” are syntactically equivalent. They are simply different modes in the genitive for expressing the same possessive relationship: YHWH’s asherah is the asherah of YHWH. Hence, Anat in Anat-YHW could be used to argue that YHWH’s asherah refers to the goddess Asherah.

Nevertheless, in the final analysis the theoretical argument that a proper name such as Asherah could carry a pronominal suffix is beset by a number of problems. First, although from a materialist perspective deities in the ancient Near East typically had properties of both common and proper nouns, they were nevertheless treated in practice as quasi-distinct persons, i.e. unitary entities. For example, within the immediate context of worship at local cult centers such as Samaria, Teman, and Jerusalem YHWH was not regarded primarily as a member of a class of deities but as the one YHWH relevant to the worshipping community. Consequently, we would not expect the discourse surrounding divine names to completely upend conventional norms of the spoken language for distinguishing common vs. proper nouns (Wiggins 1993: 188; Tropper 2001: 100). As a matter of linguistic function, the lexeme asherah cannot simultaneously inhabit both determined and indeterminate categories. If ʾšrth is correctly interpreted as the substantive asherah with an attached pronominal suffix it must not refer to the goddess Asherah. By definition the suffix distinguishes this asherah from every other asherah: this asherah is YHWH’s asherah.

Second, as Z. Zevit has noted, to specify a deity by linking it to a geographical location is not the same thing as specifying a deity by linking it to another deity through an attached pronominal suffix (2001: 402-403). The first distinguishes a deity by extending its name as it were, so YHWH of Samaria and YHWH of Teman become unique entities in their own right, whereas the second implies that the divine designation falls within an inherently indeterminate category. The asherah of YHWH in the inscriptions therefore seems to be marked as indeterminate in a way that YHWH is not.

Third, the evidence of KQom militates against interpreting the suffix on asherah as a syntactical pointer whose purpose is limited to specifying the exact cultic manifestation of the goddess. In this version of the blessing formulae YHWH appears without a geographical epithet attached to his name while the construction ʾšrth is unchanged. As YHWH is apparently being invoked here in his more mythological, non-specific form, the retention of the suffix on asherah shows that it was not unessential to conveying its intended semantic nuance.

Fourth, it is not clear that the examples of bound constructions that have been cited to demonstrate the linguistic feasibility of linking personal names of deities should be interpreted as such. Anat, Athtar, and Ashim are indeed found in bound constructions with other known deities, making the existence of a West Semitic theology of divine possession difficult to deny. But do we know that these are proper names in the sense that they designate particular divine identities? A number of scholars have theorized that because the name Anat is linked with a number of different deities and found in some contexts only in the bound form this suggests that the name has been generalized to a common noun of some kind and is not a reference to a specific deity (Dijkstra 2001b: 122; Becking 2003: 224; Day 1999: 41). B. Levine has proposed that the same may be the case with the bound forms from Ebla as well (2014b: 181). We will return to this issue below.

Finally, the theory that inscriptional asherah refers to the goddess Asherah relies on the assumption that there was no other Asherah in the pantheons of ancient Israel-Judah and that YHWH had taken the place of El in the cult. Yet traces of YHWH’s second tier status under El can still be found within the record of the HB (Römer 2015: 78-82, 127-128; Smith 2001: 142-148) and R. Thomas has argued elsewhere that during the period when the inscriptions were written the portrayal of YHWH in cultic iconography as a calf and Bes-like Horus figure suggests he was conceptualized as a younger warrior deity and son of a primary mother goddess (in press).


ʾšrth= Asheratah/Ashirtah

The second approach to interpreting ʾšrth has grown in popularity in recent years, in part because it avoids the problem of the suffix altogether. By reading the he as an element of the name, whether as a secondary feminization (Zevit 1984: 39-47; 2001: 363-366), a final frozen -a vowel (Angerstorfer 1982: 7-16; Hess 1996: 209-219; 2007: 289), or secondary lengthening of a case vowel (Tropper 2001: 81-86; cf. Naʾaman 2012: 305; Naʾaman and Lissovsky 2008: 186-208), these scholars obtain the divine name Asheratah or Ashirtah, which functions in the blessings as an independent deity parallel with YHWH. Yet despite the simplicity of this solution and the detailed philological arguments that have been brought to bear in each case, understanding ʾšrth as a single lexical item seems the least satisfactory of the various proposals that have been made to explicate its form and meaning.

Z. Zevit (1984: 39-47; 2001: 363-366) has proposed the novel idea that ʾšrth is a dialectical variant of the name Asherah and that the -h on the name reflects a process of “secondary feminization” that occurred in Hebrew early in the first millennium. According to Zevit, when the old feminine marker on nouns shifted from –at to –ah, on some nouns the new feminine marker was attached to the old. In support of this analysis, he argues that 1) the -h in ʾšrth in the inscription from KQom lacks a clear grammatical antecedent, which suggests that it does not function as the morpheme of a 3 m. sg. pronominal suffix; 2) in epigraphic Hebrew of the eighth century -h was commonly used as a mater lectionis for a long – vowel; and 3) a number of Hebrew nouns, toponyms, and personal names preserve evidence of this secondary feminization.

However, this analysis of ʾšrth is not convincing for several reasons (cf. Müller 1992: 31-32; Wiggins 1993: 170-171; Merlo 1994: 32; Renz 1995b: 92; Gogel 1998: 60-61 n. 95; Emerton 1999: 316 n. 2; Hadley 2000: 98; Day 2000: 52; Tropper 2001: 101; Schmidt 2002: 104-105; Smith 2004: 188; Rollston, 2007: 99). First, I have already noted earlier that the KQom inscription presents no real challenge to reading YHWH as the intended antecedent of the suffix on asherah. The line reads well as a simple breakup of a stereotyped phrase, so that mṣryh “from his enemies” is connected with the following imperative “save him” and the w- on mṣryh proleptically coordinates YHWH with ʾšrth: “Blessed be Uriyahu to YHWH and to his asherah, save him from his enemies….” The isolated reference to ʾšrth below and to the left of the hand symbol is also preposed with a coordinating w-, showing that it presumes an earlier invocation of YHWH. Second, that he could serve as a mater lectionis for a long -ā vowel is no argument in favor of interpreting the letter as the feminine ending –ah, since it is clear that he functioned as a mater lectionis for long -ō (or perhaps consonantal -ahu) during this period as well (Gogel 1998: 59-60). Third, the existence of secondary feminization as a linguistic phenomenon in Hebrew has not been demonstrated and the alleged examples cited by Zevit admit of more plausible explanations. For example, the ָתָה nominal ending is peculiar to poetry and P. Joüon and T. Muraoka have suggested that it performs a rhythmic function and in some cases seems to have been chosen to “avoid the contact of two stressed syllables” (2008: 259). Long ago M. Tsevat noted in his study of the Psalms that “whenever the word meaning ‘salvation’ is preceded by a ל + a substantive or a pronominal suffix, it appears in the simple form יְשועה, while whenever it precedes them, it assumes the longer form יְשועתה; this latter form is reserved solely for these cases. The same situation prevails with regard to עזרה and עזרתה, showing that this is not accidental” (1955: 21). More recently J. Tropper has explained the ָתָה ending on poetic forms and toponyms as preserving the old absolutive case, which was used for predicative expressions and in syntactic isolation (2001: 96-97). Fourth, the spontaneous addition of a feminine ending is highly implausible from a linguistic point of view. At this stage of Hebrew the tendency was to lose final vowels, not gain them (H. Gzella 2011: 438). A basic principle of historical philology is that we should expect exceptional morphological features to reflect inheritance rather than morphological addition/expansion. Finally, Zevit provides no evidence for the existence of dialectical variation in the use of a final long -ā vowel and we will see below that all known forms of Canaanite asherah can be explained on the assumption they derive from a single basic form.

Hess (1996: 209-219; 2007: 289) has observed that in Late Bronze syllabic documents the divine name Asherah is consistently spelled with a final short vowel, most often –a (=aširata). He therefore supposes that ʾšrth from KA and KQom may reflect this older spelling with the he representing a final –a vowel. But the spelling of the name Asherah in the Late Bronze Age is not a reliable guide for determining how it was articulated much later in Hebrew during the late Iron Age (ninth-eighth centuries). Although Hess notes that a final short -a vowel seems to be preserved on some feminine singular endings in Shishak’s list of Canaanite toponyms from the tenth century, during the following centuries Hebrew and other Northwest Semitic languages gradually lost short final unstressed vowels. The –atu/a/i ending became -at and eventually -at became –ā for most dialects. Because asherah is a nominal pattern with an easily identifiable feminine ending, we should assume that it experienced the same phonological development as other feminine nouns. Divine names were not a privileged category or somehow immune from general phonological change, as shown by the spelling of Asherah in the Bible.

Moreover, all known forms of asherah attested in Northwest Semitic of the first millennium can be explained within this framework of final vowel loss. The lexeme asherah at KA can be assumed to have been vocalized with a final –at, since the building complex seems to have been built and used by Israelites from the Northern Kingdom and some evidence suggests that Israelite Hebrew retained final –at in the feminine singular (Garr 1985: 59-60, 93-94; Gogel 1998: 188 n. 232; Na’aman and Lissovsky 2008: 199-200, n. 9). Assuming that this is correct and that the -h represented a suffix morpheme, ʾšrth would have been pronounced ʾašerātō or even ʾašerātahu (Gogel 1998: 60-61 n. 95). The inscription lʾšrt from Tel Miqne/Ekron also preserves a spelling of asherah with a final –at, thus ʾašerat. In a number of respects the language used for inscriptions at Ekron is close to Phoenician and northern Hebrew, so the preservation of the feminine ending –at is explicable (Gitin, Dothan, and Naveh 1997: 8-15). On the other hand, even though ʾšrth at KQom has the exact same orthography as ʾšrth at KA, because the inscription was written in Judah it is conceivable that asherah by this period was pronounced with a final –ā rather than -at. As C. Rollston has observed, the conversion of final –ā to –at is the “standard development that occurs when a tertia he noun has a pronominal suffix” (2007: 99). When a pronominal suffix is added, ʾašerah > ʾašerat > ʾašerātō. Finally, the ʾašerah of biblical Hebrew presents the same –ā feminine ending, the end result of the loss of case vowels and final –at in the absolute.

Coming from a very different angle, J. Tropper (2001) has argued that the final he in the divine name YHWH reflects a spelling with the old absolutive case –a, a case vowel common to early West Semitic proper names and preserved sporadically in Hebrew of the first millennium. Because the name YHWH appears to preserve this old absolutive case, he proposes that the he on ʾšrth at KA and KQom should be analyzed the same way, with the he representing the case marker –a that has been secondarily lengthened. However, this explanation of ʾšrth faces a number of challenges as well. While Tropper has made a compelling argument for a final –a vowel on the long form of the name YHWH and for the preservation of the absolutive case on various word forms and toponyms, not all of his proposed examples are convincing. For example, the significance of masculine personal names appended with a final he or aleph is unclear. Traditionally understood as hypocoristics with a vocative ending, it is not difficult to imagine that such a function could have outlived earlier usages of the case vowel on proper names (Noth 1928: 38). It is also unclear whether the exclamatory particles extended with an –a ending should be understood on the same level as the –a ending on proper names (cf. GKC: 471). Finally, the use of the expression dwdh in the Mesha inscription to support the thesis that divine names during this period often featured an –a case ending is questionable. We lack clear evidence that a god Dwd existed and was worshipped in the southern Levant. The “DRK of Beersheba” in Amos 8:14 makes good sense as an epithet of a local form of YHWH, “the Strength of Beersheba” (cf. Blenkinsopp 2003: 161) and the syntax of the phrase ʾrʾl dwdh, which is apparently a construct relation, would be more intelligible if we assumed the first component ʾrʾl were a reference to a deity/cult statue and the last component with an attached pronominal suffix something else.[xxv]

Further, even if we were to agree that the final –a vowel on YHWH preserves a remnant of the absolutive case, we have little reason to think that the vowel should be found on other divine names in Old Hebrew. As Tropper himself admits, the absolutive -a vowel on YHWH, if that is what it is, has been lengthened so that it is no longer a simple case vowel, i.e. it has become an integral part of the articulation of the name. Thus in a very real sense the vowel represents only a fossil related to the unique history of the divine name. In addition, the phonological situations of the endings on the names YHWH and ʾašerat are very different, with a final vowel in the case of YHWH and the f. sg. morpheme in the case of ʾašerat.

Taken together with the evidence of general phonological development in Northwest Semitic discussed above, it seems very unlikely that the archaic form Aširata would have survived to experience a lengthening of the final vowel. Rather, the final short vowel on asherah should have been reduced, in line with the loss of case vowels on other feminine nouns. There is no need to posit a new, previously unattested form of the divine name Asherah in Hebrew that would be a complete outlier in terms of what we would expect linguistically.

Beyond linguistic considerations, the interpretation of the –h on ʾšrth as a suffix morpheme should be preferred on contextual grounds. Because ʾšrth is associated with both YHWH of Samaria and YHWH of Teman in the blessings from KA, the cultic context would seem to necessitate distinguishing one ʾašerat from the other. It hardly makes sense to speak of two local forms of YHWH and only one goddess Asherata as their partner! In addition, we have already noted the comparative data for the widespread cultic convention of divine possession in West Semitic. Female deities were often marked as belonging to their male partners. So the construction “YHWH of GN and his asherah” is acceptable from a religio-historical perspective.



The third proposal has received less attention in scholarly literature, most likely because it is not as straightforward as the other interpretations and has been regarded as overly speculative. This approach recognizes that asherah likely represents a deity with an attached pronominal suffix, but goes further than the first alternative by positing a different semantic valence to the term. On this view asherah was originally a divine name of a particular goddess that has been generalized (Pardee 1995: 301-303; 2005: 281-285; Levine 2014a: 171-91; 2014b: 39) or was fundamentally a common noun that could be used as a proper name or a generic title depending on the context (Margalit 1990: 264-297; Binger 1997: 145-147; Wyatt 1999: 104). In either case, inscriptional asherah is thought to mean something like “divine consort/asherah goddess.”

Although the different variants of this approach are not equally persuasive or have relied on questionable arguments, the general line of interpretation of postulating a common noun meaning to asherah is appealing as a solution to the problem of ʾšrth. After all, this understanding of asherah fits the semantic-syntactic context, which methodologically speaking should be the primary criterion for determining the usage and implication of a word in an unfamiliar linguistic setting. We have already seen earlier that asherah cannot have reference to the proper name Asherah because of the attached pronominal suffix, but must refer to a common noun of some kind, an asherah differentiated from all other asherahs. Not surprisingly, Hebrew epigraphists have often felt constrained to gloss inscriptional ʾšrth as “his consort” (Gogel 1998: 60; cf. DNWSI 1:129; McCarter 1987: 147).

As indicated above, there are two possible ways we could explain how asherah came to have a common noun usage in Late Iron Age Israel-Judah, the first that the divine name Asherah itself originated as a common noun and that this usage continued into the first millennium, or second that Asherah had semantically developed into a generic title in the context of Israel-Judah.

The first option is at least theoretically plausible. Many divine names in West Semitic appear to have originated as common nouns or had a common noun valence, so it would not be unreasonable to suppose that the same might be the case with Asherah. For example, El, Baal, Elat, Rahmay are common nouns used as if they were proper names: El= (the) God, Baal= (the) Lord, Elat= (the) Goddess, Rahmay= (the) Womb. In Ugaritic Asherah is often invoked parallel to common noun originating epithets, such as Elat and Rahmay (KTU 1.14 IV 34-39; 1.3 V 36-37; 1.15 III 25-26; 1.23.13, 28). In addition, the plural ʾašērōt is used in the HB to designate a class of female divinities associated with various Baal deities (Jdgs 3:7), the same as ʿaštārōṯ (Jdgs 2:13; 10:6; 1 Sam 7:4; 12:10). That the term asherah had a generic/titular function would also shed light on how it came to be applied to multiple goddesses geographically and chronologically separated from one another (Binger 1997: 146-47).

So what would asherah have meant if it were originally a common noun? B. Margalit (1990: 264-97) has proposed that asherah is a long forgotten Northwest Semitic noun derived from the root ʾṯr, “to follow,” and denotes “wife, consort.” In support of this etymology, he notes that the noun aṯt “wife” is invoked parallel to aṯrt: ks . qdš ltphnh . aṯt . krpn ltʿn . aṯrt “a holy cup a wife may not see, a goblet even an aṯrt may not see” (KTU 1.3 I 13-15), and once in a letter aṯrt apparently bears an attached pronominal suffix, laṯrty “for my asherah” (KTU 2.31:42). Further, he correlates this proposed etymology with the drawing on pithos A from KA depicting YHWH and “his asherah” standing behind and argues that the understanding of asherah as “one who follows” or “wife” is reflected in the biblical metaphoric usage of “to follow/go after” as an expression for marital fidelity.

However, this analysis meets with a number of challenges. First, the meaning of asherah as “wife, consort” is not clearly attested anywhere in Semitic. From the earliest of times asherah seems to have been consistently used as a proper name, including in Ugaritic, with anomalous usages showing up only later in Hebrew (Ebeling 1928, 1:169; Perlman 1978: 73-78; Day 1986: 385-408; Wiggins 1993: 132-64, 175, 192). It is difficult to believe that the term could have originated as a common noun and yet leave no clear attestation of this usage preserved in Semitic more generally. Second, the analysis of asherah as a fem. active participle form from the root ʾṯr is unlikely. The name Asherah is never found in a compound phrase such as other participle forms used as divine epithets (e.g. bny bnwt, qnyt ʾilm), and it is doubtful that the verbal notion of “following” in itself would be sufficient to contain the concept of wifehood. Margalit neglects to notice that the verb ʾṯr “to follow” is never associated with wives or consorts in Ugaritic (DUL: 126). With regard to a participial explication of aṯrt, A. Rahmouni has commented, “the process by which such a participle, originally allegedly used as a compound noun in a divine epithet, eventually became a divine name in its own right, still awaits clear precedent,” and further, if the name Asherah were formed from a participle we would expect it to be vocalized in Hebrew “in accordance with that of the active participle of action verbs, namely * ʾōšeret or possibly ʾōšərā, for neither of which is there any evidence” (2008: 283). Third, the mention of aṯrt parallel to aṯt “woman, wife” in KTU 1.3 I 13-15 need not be understood as synonymous parallelism. As argued by S. J. Park, “It would be better to understand the second clause in an emphatic sense, ‘a sacred cup woman may not see || a goblet even Athirat may not see’” (2010: 532). Fourth, the alleged use of aṯrt with a pronominal suffix appears in a broken context and it is unclear whether the y is actually a suffix or is rather the beginning of another word. Several scholars have reconstructed an original aṯrtym with no word divider (Heide 2002: 110-11 n. 2). The inscription is simply too fragmentary to use as a basis for positing a common noun meaning to aṯrt in Ugaritic. Finally, while the language hālak ʾḥarê “to go after” in Hebrew seems to be closely linked to the conceptual domain of marital fidelity/amorous relations, the expression has no discernible relation to the root ʾšr “to follow,” and it is methodologically dubious to try to explicate the imagery of pithos A on the basis of a presumed etymology (Keel and Uehlinger 1998: 219, n. 50). The concept of a wife “following after” her husband could have existed and even lay somewhere in the background of the portrayal of YHWH and his asherah on the pithos and yet have no relation to the etymology of asherah.

Binger (1997: 146-47) has similarly suggested that asherah was originally a secular name-title, whose common noun status would have allowed it to apply to many different goddesses. She speculates that the title was generally used to designate female counterparts of high gods such as El, Baal, and YHWH. But unfortunately Binger fails to provide any solid evidence for this original common noun meaning or to explain why asherah in the extant texts is used consistently as the proper name of a goddess. Furthermore, it is conceivable that the divine name Asherah had a quasi-titular function in the Late Bronze-Iron Age southern Levant (outside Israel-Judah) and yet was not in fact understood as a regular common noun that could be declined with a suffix. We will see below that the same may be the case with the divine name Astarte during the first millennium BCE.

In sum, we lack sufficient evidence to support the notion that asherah originated as a common noun. All extant material suggests it was used as a divine name throughout the bulk of its millennia long history. Although aṯrt is sometimes used parallel with other common noun originating epithets in Ugaritic, these epithets are nevertheless treated as proper name designations, so their relevance to understanding the meaning and origin of asherah is moot.

The only remaining alternative is to assume that the name Asherah had semantically developed into a common noun as an innovation peculiar to Israel-Judah or the central southern Levant. B. Levine (2014: 181) has proposed that asherah may have been generalized into a synonym for ʾilt “goddess,” comparable to the development of Ishtar in Akkadian into a general term for “goddess” (e.g. ištaru, ištartu, etc.), and D. Pardee has similarly assumed that the term asherah as found in the inscriptions from KA and KQom represents the end product of a long process of semantic evolution whereby the word came to have the common noun meaning “consort” (1995: 301-303; 2005: 281-285).

In fact, we may have specific evidence that this semantic development had indeed occurred. As I mentioned above, the only place that we find anomalous usages of the term asherah in Northwest Semitic is in the context of Israel-Judah. First, the plural ʾašērōt appears in Jdgs 3:7 as a designation for a class of female deities worshipped in association with various Baals. The use of asherah in this manner suggests that for the biblical authors the term had a generic valence of some kind. Although ʾašērōt here has often been thought to be a scribal mistake for ʿaštārōṯ, because the latter are associated with Baals elsewhere (Jdgs 2:13; 10:6; 1 Sam 7:3, 4; 12:10) and the reading ʿaštārōṯ is reflected in the Syriac and Vulgate (Wiggins 1993: 102; Hadley 2000: 63-64), the earliest versional evidence in the Septuagint supports the MT reading. Day has correctly noted that ʾašērōt is the lectio difficilior and it is much “easier to understand how ‘the Asheroth’ could have become corrupted here to ‘the Ashtaroth’ than the other way around” (Day 2000: 45). The Israelite Asherah is also closely associated with Baal later in the Dtr narrative (Jdgs 6:25-26; 1 Kgs 16:31-32; 2 Kgs 17:16; 21:3-7; 23:5-6), so the appearance of the divine pair here is explicable at a narrative level. By referencing the Baalim and Asherot (Israelite) instead of the Baalim and the Ashtaroth (Phoenician or non-Israelite), the Dtr author seems to be trying to link the former with the latter and stigmatize them as foreign/non-Israelite (cf. Olyan 1988: 10 n. 28).

Second, E. Larocca-Pitts has observed that the term ʾašērîm in the Bible appears to refer to female cult statuary and not the statuary of Asherah in particular. The biblical authors enjoin or record the destruction of the ʾašērîm of many female deities, including Phoenician Astarte in Jerusalem (2001: 191). If this understanding is correct, then it suggests that the term ʾašerah had developed a common noun valence that allowed it to be generalized to other female divinities. That is to say, the biblical condemnation of the ʾašērîm presupposes a semantic evolution from Asherah as proper name > to asherah as common noun for multiple goddesses, and then > to asherah as term for female cult statuary.

Third, the generic quality of the term asherah is also suggested by the consistent attachment of the article to it when used by the biblical authors to designate the goddess Asherah (1 Kgs 15:13; 18:19; 2 Kgs 21:7; 23:4). Asherah is used as if it were a common noun title comparable to Baal and was in need of further determination: hāʾašērâ “the asherah” implies that she is a particular member of a larger class of ʾašērōt.

Although it is unclear what exact nuance may have attached to asherah at KA and KQom, a hitherto unknown common noun meaning peculiar to the culture of Israel-Judah seems the most workable of the current alternatives for interpreting the phrase ʾšrth. We already mentioned above that semantic development of a divine name is plausible in a comparative context (e.g. Ishtar). From the little we know, divine names in the ancient Near East seem to have had extraordinarily complicated histories. Not only do many divine names appear to have originated as common nouns, but their meaning could change and evolve over time, depending on the historical and cultural circumstances.

As a point of comparison, we can note two other names of goddesses in the Syro-Palestinian region whose meaning seems to have experienced semantic development from the Late Bronze to the Late Iron Age or at least show semantic divergence in their interpretation by particular cultures. Both Astarte and Anat appear in second millennium sources from Ugarit as youthful, second-tier warrior goddesses. As members of the household of El, they are conjugally unattached to any males and non-reproductive in character (Schmitt 2013: 217; Smith 2001: 54-66; Day 1999: 36-43; Wyatt 1999: 109-114; Pardee 2007: 27-39).[xxvi] Anat is regularly referred to by the epithet btlt, indicating her status as a young female who has not born children (Smith 1994: 8-9 n. 20). Scattered evidence suggests that these profiles of Astarte and Anat as young martial deities may have persisted on into the first millennium BCE in some quarters.[xxvii] Yet by the ninth-seventh centuries BCE the divine names Astarte and Anat seem to have been used to designate very different types of deities in respectively Phoenician and Aramaean contexts. In Phoenicia Astarte has become a first-tier primary goddess, functioning as wife and mother. At Tyre, Sidon, and elsewhere she is regularly given the epithet rbt, reflecting her status as chief female in the pantheons and mother of the divine royal heir (Gordon 1988: 127-132; Smith 2002: 129; Smith 2006: 100-101).[xxviii] At Sidon she is linked to the Baal of Sidon and the youthful deity Eshmun in a divine triad, the epithet “Name of Baal” identifying her as the consort of Baal (Schmitt 2013: 217).[xxix] Although the epithet “Name of Baal” is identical to that of her predecessor at Ugarit, it would be a mistake to assume that she is a direct continuation of the earlier goddess. The earlier strong martial connotation of the epithet is no longer in evidence and it now seems to function as a Sidonian equivalent to the Punic epithet “Face of Baal” used for Tinnit, pointing to her role as a mediator with the people to her husband (Garbati 2013: 531-32; Seow 1999: 322-325). Theophoric personal names suggest she was conceptualized as a benevolent figure (Benz 1972: 386-87; Smith 2002: 129; Schmitt 2013: 217; Abousamra and Lemaire 2013: 153-157) and R. Schmitt has noted that “in the first millennium glyptic repertoire from Phoenicia the motif of the armed goddess is completely absent” (2013: 217).[xxx]

Moreover, a variety of data indicate that Astarte was closely related to and sometimes identified with Tinnit, the Punic first-tier primary goddess and incarnation of Canaanite Asherah. This includes the compound name tntʿštrt on the Sarepta plaque, the supplanting of Astarte by Tinnit as chief goddess at Carthage and concurrent use of Astarte theophorics in personal names, and the association of Astarte at Carthage with “the Tinnit in Lebanon” (KAI 81).[xxxi] The latter inscription has often been thought to show that Astarte and Tinnit were understood to be separate deities in the Punic world (Cross 1973: 29-30; Olyan 1988: 54-56). However, the specification of Tinnit as “the Tinnit in Lebanon [i.e. Phoenicia proper]” implies that Astarte is being differentiated only from that particular form of Tinnit. Indeed, the association of Astarte with Phoenician Tinnit suggests that Astarte may have been a name or title of the Carthaginian Tinnit.[xxxii] Astarte’s motherly status is further indicated by her frequent identification with several other mother-type goddesses in the Mediterranean world, including Isis, the goddess of Tas Silg, Uni/Juno, and the Baalat of Byblos.[xxxiii] In Philo‘s Phoenician history Astarte is remembered as the wife of Kronos and Zeus and mother of various deities (Attridge and Oden 1981: 51, 53, 55).

On the other hand, in Aramaean cultural contexts Anat has become a regular name of the wife of the primary high god. In Essarhaddon’s treaty with Baal of Tyre Anat is paired with Bethel in the form Anat-Bethel. Because Bethel was an Aramaean form of the high god El, Anat-Bethel likely refers to the goddess functioning as his first-tier wife (van der Toorn 1992: 83-85; Röllig 1999: 174). Similarly, Anat appears in Upper Egypt paired with Yahu/Bethel in contexts that are strongly suggestive of her domestic wifely status. As was mentioned above, the goddess is found in a familial triad connected to Yahu’s temple at Elephantine where Yahu is implied to be father, Anat-Bethel his wife, and Eshem-Bethel their son. In Syene Anat probably lies behind the “Queen of Heaven,” who is said to have had a temple there associated with the temple of Bethel (van der Toorn 1992: 97; Porten 1968: 164-65). In the Amherst Papyrus (Steiner 2003) the wife of Bethel, conventionally referred to as Marah (“the Lady”), is described as a heavenly goddess and once called Anat (col. VII.7-19). Based on the motifs associated with her, she is clearly a mother-type goddess: 1) a frequent refrain in the text is that she reared, suckled, and nourished the people as her children (I.17-19; II.14-III.6; III.14); 2) just as Bethel is called Bull and Father, she is symbolized as the divine Cow (II.14-III.6; XIII.1-9; XVI.7-19); and 3) she is a goddess of sacred marriage: lying in the “waters of fertility” she eternally brings forth fruit to her consort (II.8-11; XVI.7-19).

All of this data is difficult to explain unless we assume that the divine names Astarte and Anat had developed in Phoenician and Aramaean cultural contexts so that they no longer had reference to the deities who bore those names in the second millennium. Over time the names had been given new and decisively different semantic content. Instead of the independent and martial second-tier goddesses encountered at Ugarit, Astarte and Anat had become permanently wedded, motherly in character, and primary in their respective pantheons.

In fact, Astarte and Anat seem to have been regularly used as generic titles during the first millennium. The name Astarte was applied to many different local goddesses during the Persian and Hellenistic periods, including goddesses from Sidon, Tyre, Byblos, and Ashkelon in the southern Levant (Fuks 2000: 27-48; Stager 2005: 427-49), and often is found in apposition to other divine names or accompanied by descriptive epithets, suggesting that it had a mythological nuance capable of further definition.[xxxiv] The multiplication of Astarte deities especially in the late period explains why the authors of the Dtr History treat Astarte as a generic title for non-Israelite goddesses paired with local Baals in neighboring cultures (Jdgs 2:13; 1 Sam 7:3-4; 12:10). In the case of Anat, the occurrence of the name in the Amherst papyrus implies that it functioned as an epithet and not as the primary name of the wife of Bethel. Throughout the text the goddess is repeatedly referred to as Marah or Nanai, while the name Anat occurs only once in parallel to Mami, a generic term for the mother goddess (VII.7-19). Because Anat and Mami only appear in the text here, their parallel usage suggests that Anat may have been a title with connotations of motherhood and divine primacy, which would help explain why it was so easily adopted by Elephantine Jews as a designation for their chief goddess.

Alternatively, it is possible that the name Anat originated in early Northwest Semitic as a generic designation for a female sexual partner/wife and that its usage among Aramaeans preserves this earlier second millennium understanding. A plausible etymology of Anat is that it is related to ʿn “furrow,” a root attested in Ugaritic and Hebrew (cf. Deem 1978: 25-30; Walls 1992: 114-15; DUL: 169; Smith and Pitard 2009: 149-50). The conceptualization of women as the fertile ground in which seed is planted was widespread in the eastern Mediterranean, and an understanding of Anat as denoting “the female furrow” would explain how it could be applied to various female deities and declined with a pronominal suffix. Significantly, the divine name Anat appears in second millennium toponyms of central Palestine along with Canaanite Baal and Baalat, further reinforcing the impression that it is a descriptive epithet, perhaps denoting the same goddess as Baalat (cf. Zevit 2001: 603).




By analyzing the various proposals that have been made for elucidating ʾšrth from KA and KQom we have been able to determine that it most likely has reference to a common noun denoting YHWH’s female partner: “his asherah.” This explication of the phrase not only does no violence to the evidence that inscriptional asherah is a female deity paired with YHWH, but it also harmonizes best with the lexical-syntactic evidence that asherah here is declined with a pronominal suffix with YHWH as the antecedent.

While we lack sufficient information to clarify the precise meaning of asherah in the inscriptions, the identification of the term as a common noun has important religio-historical implications and may point toward a possible answer. We have already seen above that because asherah in the inscriptions is declined with a pronominal suffix, it cannot refer to the goddess Asherah. This asherah is by definition distinguished from all other asherahs, including perhaps especially the goddess whose proper name was Asherah. Because the Bible shows that Israelite Asherah was worshipped together with Baal (=El) until the end of the monarchy (e.g. 2 Kgs 23:4), it seems possible to infer that in ancient Israel-Judah there were two goddesses concurrently designated asherah, the goddess named Asherah in the sense of a proper name (the Asherah) and the goddess that appears in our inscriptions as YHWH’s asherah. We mentioned earlier that there is some evidence that by the time of KA and KQom YHWH was still conceptualized as a second-tier deity under El and Asherah. If this was the case, then it is possible that the common noun meaning of asherah was developed to distinguish YHWH’s female partner from the Asherah associated with El.

At any rate, whatever the relationship of YHWH’s asherah to Asherah, what seems most certain is that the worship of this goddess was a general feature of Israelite-Judahite religion, as it is attested not only for YHWH cult in the south at Teman-Edom, but also for Samaria and Judah. The distribution of the forms of YHWH invoked in the inscriptions suggest a broad cultural sphere where YHWH was believed to be wedded. Furthermore, YHWH’s asherah appears to have been at home in public national cult as well as personal family religion. The Samaria and Teman mentioned in the blessings are clearly regional cult centers associated with state polities, where particular manifestations of YHWH and his asherah were related to established territorial boundaries, whereas the YHWH and asherah at KQom are expressed in the context of aspirations for the well being of an individual after death. The stereotyped nature of the blessings and KA’s links with Northern Kingdom officialdom reflect the traditional, prevailing, and authorized character of this polytheistic faith (cf. Dijkstra 2001a: 30). Finally, YHWH and his asherah were likely represented with icons or cult statues at the abovementioned sites. YHWH of Samaria and his asherah are treated in the blessings as conceptually distinct from YHWH of Teman and his asherah, and it is difficult to imagine how this distinction could have been maintained unless they were physically embodied in their respective sanctuaries.


[i] For history of scholarship, see Hadley 2000: 11-37; Wyse-Rhodes 2014: 71-90.

[ii] Inscription 3.1, Aḥituv, Eshel, and Meshel 2012: 87-91. Cf. Renz 1995a: 61; Zevit 2001: 390-392; Dobbs-Allsopp, Roberts, Seow, and Whitaker 2005: 289-292; Naʾaman 2011: 302.

[iii] Inscription 3.6, Aḥituv, Eshel, and Meshel 2012: 95-97. Cf. Renz 1995a: 62; Zevit 2001: 394-397; Dobbs-Allsopp, Roberts, Seow, and Whitaker 2005: 293-294; Naʾaman 2011: 303.

[iv] Inscription 3.9, Aḥituv, Eshel, and Meshel 2012: 98-100. Cf. Renz 1995a: 64; Zevit 2001: 398-400; Dobbs-Allsopp, Roberts, Seow, and Whitaker 2005: 295-297; Naʾaman 2011: 306.

[v] Aḥituv, Eshel, and Meshel read kl ʾs̆r. ys̆ʾl. mʾs̆. ḥnn hʾ “all that he requests from a man, that man will give him generously.” But from the photos it is difficult to make out what follows ys̆ʾl. (cf. McCarter 2003:172 n. 3) and ḥnn hʾ means “he is gracious,” not “he will give” (Naʾaman 2011: 307). Also, the sense of the line yielded by this reading is unsatisfactory. First, based on the mythopoetic focus and cultic function of the inscriptions at KA, it seems doubtful that an Israelite receiving something from an unspecified gracious man would be the subject of this short verse. Second, line 2 shows evidence of parallelism with the occurrence of “ask”//“entreat” and “gracious”//“give.” Because the adjective “gracious” is prominently associated with YHWH in the Bible (e.g. Ex. 34:5; Ps. 116:5), a more plausible reading is that ḥnn hʾ “he is gracious” refers to YHWH as the one who responds to requests.

[vi] Inscription 4.1.1, Aḥituv, Eshel, and Meshel 2012: 105-107. Cf. Renz 1995a: 58; Zevit 2001: 373-374; Dobbs-Allsopp, Roberts, Seow, and Whitaker 2005: 285-286; Naʾaman 2011: 308-09.

[vii] Inscription from KQom, cf. Renz 1995a: 207-210; Zevit 2001: 359-70; Dobbs-Allsopp, Roberts, Seow, and Whitaker 2005: 408-414; and Aḥituv, 2008: 220-224.

[viii] I assume that the placement of mṣryh after the coordinating waw and before lʾs̆rth is a poetic breakup of a stereotyped phrase and therefore that asherah functions as a second object in the blessing formula. Cf. Hadley 2000: 96-100; Naveh 2001: 194-97; Aḥituv 2008: 221-224; Schniedewind, 2013: 113.

[ix] E.g. G. Ahlström, W. Dever, D. N. Freedman, M. Coogan, H. P. Müller, C. Uehlinger, M. Dietrich, O. Loretz, C. Frevel, P. Xella, A. F. Rainey, K. van der Toorn, B. Schmidt, M. Leuenberger, C. Rollston, A. Berlejung, T. Römer.

[x] E.g. A. Angerstorfer, Z. Zevit, M. O’Conner, D. Conrad, R. Hess, S. B. Parker, N. Na’aman.

[xi] E.g. Z. Meshel, M. Pope, B. Margalit, D. Pardee, B. Halpern, T. Binger, N. Wyatt, B. A. Levine.

[xii] The cult object interpretation was first proposed by Lemaire 1977: 595-608 with reference to Khirbet el-Qom and soon thereafter was applied to the inscriptions from Kuntillet ‘Ajrud.

[xiii] See Deut 7:5; 12:3; 16:21; Jdgs 3:7; 6:25, 26, 28, 30; 1 Kgs 14:15, 23; 15:13; 16:33; 18:19; 2 Kgs 13:6; 17:10, 16; 18:4; 21:3, 7; 23:4, 6, 7, 14, 15. Outside of the Dtr historical narrative of the premonarchic and monarchic periods, asherah occurs in Ex 34:13, a chapter commonly viewed as having strong Dtr connections, in some late literary/redactional (Dtr?) strata in prophetic narrative (Isa 17:8; 27:9; Jer 17:2; Mic 5:3), and in the synchronistic references in Chronicles.

[xiv] On the lack of correspondence between theophorics contained in personal names and the full gamut of deities worshiped in particular Syro-Palestinian cultures, see Smith 2002: 4-5; Pardee 1988: 119-51.

[xv] See also KAI 104; 105; 175; Constantine N42; Tirekbine N1. The singular emphasis on Baal in Punic dedications has been noted by Garbati 2013: 530; Amadasi Guzzo and López 2012-13: 176 n. 80.

[xvi] For discussion of the relationship between deities and their cult images, cf. Jacobson 1987: 15-32; Collins 2005: 33-35; Dick 2005: 51-58; Hundley 2013: 277-81; Pongratz-Leisten and Sonik 2015: 3-69.

[xvii] I follow Porten’s reconstruction (1968: 317) of the missing text except restore YHW rather than Herem, as I explain later below.

[xviii] For the identification of Teman with the core territory of the Edomite state, see Edelman 1995: 10-11; Knauf 1992: 347-48.

[xix] Allen 2011 has expressed skepticism that geographical names in the construction DN b-GN are integral to the identity of the deities so construed on the basis that they never appear in embedded god-lists or contrast two deities of the same first name. However, this methodological criteria for detecting divine individuation is too exacting and simplistic to be very useful, as god-lists are very unusual in West Semitic inscriptions. Although Allen is correct that many of the known cases of DN b-GN do not function precisely as full names in the same manner as Ishtar of Ninevah or Ishtar of Arbela, his analysis fails to recognize that divine nomenclature was generally more fluid based on the rhetorical and discourse context and a deity did not always or consistently have to be specified by a geographical name if its identity was already clear in context. In any case, most of the occurrences of geographical names in DN b-GN seem to modify the name of a deity or at least localize its manifestation (e.g. Ps 65:2; 99:2; 1 Sam 5:5; 2 Sam 15:7; KAI 17:2; 81:1; 181:13; Lapethos 6; Ammonite seal). Also, the interchange of expressions such as mlk ʿṯtrt “Mlk of ʿAṯtrt,” mlk bʿṯtrt “Mlk in ʿAṯtrt,” and mlk… yṯb bʿṯtrt “Mlk… who resides in ʿAṯtrt” at Ugarit and yhw ʾlhʾ byb “Yahu the god in Yeb,” yhw ʾlhʾ (zy) byb brtʾ “Yahu the god in Yeb the fortress,” and yhw ʾlhʾ škn yb brtʾ “Yahu the god who dwells at Yeb the fortress” at Elephantine point to their commonality in function parallel to the formula DN-of-GN. Cf. McCarter 1987: 140-141; Niehr 2003: 194; Sommer 2009: 38-39; Smith 2012: 214-215; Hutton 2012: 183-187.

[xx] אשם/אשימא (with prosthetic aleph) is attested as an epithet of a male deity in a dedicatory inscription from Teima and in the compound divine name Ashim-Bethel at Elephantine, Niehr 2014: 169-170.

[xxi] In the past the grammatical model that has been used to describe this phenomena is “double determination,” but this seems to reflect a modern conceptual framework that understands divinity to be essentially a determinate category.

[xxii] Although Pardee (1995: 301-303) has proposed reading the –h as an adverbial/locative morpheme, he provides no clear parallel for this obscure usage. The fact that “his Anat” has an immediate antecedent in Gatharu and elsewhere Anat is associated with the same deity lends support to understanding the –h as a suffix. Cf. Smith 2001: 72-74; Xella 2001: 74-75; Schmidt 2002: 104-107; Heide 2002: 110-11 n. 2.

[xxiii] According to von Pilgrim 1999: 142-45, the temple of Yahu was founded in the sixth century during the 26th dynasty, which lends support to the assumption that Judeans had already been living in Elephantine for a space of time.

[xxiv] P. Grelot rejected both of these restorations and followed the earlier suggestion that the line continues the patronymic with br X… (1972: 95). But this reading can be excluded based on 1) the presence of a conjunctive w- before Anat-YHW, which shows that another divine name was listed in the gap; 2) the preservation of a final aleph after the break, which fits well with a restoration of the epithet אלהא; and 3) the parallel oath in AP no. 6/TAD B 2.2 that suggests a divine name should fall before the geographic localization.

[xxv] The term ḏd in Ugaritic appears to designate a tent-sanctuary, cf. Clifford 1971: 221-27; Watson 1995: 221-222; DUL: 285.

[xxvi] Similar to Ishtar, Astarte and Anat in the Ugaritic texts are symbols of potential sexuality rather than its realization through motherhood.

[xxvii] The martial profile of Astarte seems to have had greater longevity in Egypt, cf. Schmitt 2013: 217. Anat was identified with Athena on Cyprus and presumably lies behind the reference to Athena in Philo of Byblos, Day 1999: 39.

[xxviii] Cf. Rahmouni 2008: 281-82, who supposes that rbt was an honorary title that could be applied to many different goddesses (e.g. Šapšu). But the epithet does not seem to have been used haphazardly in the various pantheons in which it occurs; generally only a single goddess receives the epithet, and a number of scholars have hypothesized that Šapšu is closely related to ʾAṯiratu and perhaps even an avatar.

[xxix] Although Eshmun is never explicitly designated the son of Baal and Astarte, his identification as קדש שר “holy prince” (KAI) and interpretatio graeca as Apollo underlines his youthful second-tier character.

[xxx] For Phoenician Astarte’s ambiguous iconography, see E. Bloch-Smith 2014: 167-94.

[xxxi] On the relationship between Astarte and Tinnit, cf. Bonnet 1991: 82-83; Lipiński 1995: 203-04; Amadasi Guzzo 2004: 47-54; Peckham 2015: 305, 393; 542-557. Cf also the use of Tinnit’s epithet “Face of Baal” for Astarte at Ashkelon, Stager 2005: 427-449.

[xxxii] The name Tinnit may have been a kind of generic title as well, derived from the root MNY/MNH “to count” and related to good fortune and the passage of time, Vennemann 2008: 569-90.

[xxxiii] For Isis, see Abousamra and Lemaire 2013: 156-57; the goddess at Tas Silg, Amadasi Guzzo 2010: 465-89; Uni/Juno (KAI 277); and the Baalat of Byblos, Bordreuil 1998: 1153-1164; Zernecke 2013: 226-242. Cf. Bloch-Smith 2014: 167-94.

[xxxiv][xxxiv] Cf. Astarte of the Mighty Heavens (KAI 14); Astarte-Hor (KAI 294); Isis-Astarte (KAI 48); Tinnit-Astarte (KAI 285); Astarte-Eryx (CIS I 135.1); Astarte-Kition (KAI 37); Astarte-Malta (Bloch-Smith 2014: 181); Astarte-Paphos (RES 921.3-4); Aphrodite Urania (Herodotus 1:105); Astarte Megiste (Philo