New Evidence for Asherata/Asherah

Abstract

This paper examines the appearance of published West Semitic spellings of the name of the deity commonly referred to as Asherah. In light of new evidence from the Bronze Age Amorite sources, as well as the complete publication of the inscriptions at Kuntillet ʿAjrud, a review of the analysis and discussion concerning the identification of the deity is undertaken. The purpose will be to ascertain the significance of the witness of epigraphic Hebrew texts at Kuntillet ʿAjrud and Khirbet el-Qom in light of earlier Bronze Age evidence, the biblical attestations, the conceptualization of deity, and the understanding of Iron Age epigraphic Hebrew spellings of the feminine singular suffix, as well as pronominal suffixes. The more complete availability of textual witnesses provides a foundation on which to argue the degree of continuity across more than a thousand years of the appearance of the deity in the West Semitic world.

Some years ago, this author attempted to catalog all published syllabic and alphabetic spellings of the goddess commonly referred to as Asherah, as the spellings appeared in West Semitic texts of the Bronze and Iron Ages.1 While attestation of the goddess’ names does occur in later East Semitic and Hittite contexts, the focus here will be on Old Babylonian and West Semitic and recent publications.2 In the earlier publication, I argued that the original form of the divine name was Asherata and that later developments of the name in West Semitic could be explained as developments from this spelling. The question arises as to whether this argument remains cogent in light of recent discoveries and texts bearing this West Semitic name, and whether the spelling has a masculine singular suffix, as argued for its appearance in inscriptional materials such as those found at Kuntillet ʿAjrud. Despite these arguments, some scholars continue to assume, without further discussion, that the ʾšrth following yhwh and the conjunction should be interpreted as “his Asherah”.3 The evidence will support the absence of this pronominal suffix.

1. Bronze Age Evidence

In the absence of any earlier attestations at Ebla or elsewhere, the first occurrences are found in Old Babylonian.4 Twice there appears the Amorite personal name spelled ^daš-ra-tum-mi, with the sense “Ashratum is my mother”.5 The divine name also appears in three Babylonian seal impressions, two with the spelling of ^daš-ra-tum and a third appearing as ^da-ši-ra-tum.6

The name occurs in late second millennium BC West Semitic archives only at Ugarit and Amarna. There are no attested occurrences elsewhere in the published texts of the Late Bronze Age West Semitic world. In addition, the complete reading and analysis of all personal names at Alalakh Levels IV and VII results in some 7971 name attestations.7 Not only is the divine name of the goddess unattested at Alalakh, there are no examples of its appearance in personal names.8

At Amarna and Ugarit, the evidence is different. The name of the ruler of Amurru in the mid-fourteenth century BCE, abdi-a-ši-ir-te, occurs ninety-five times in sixty-two Amarna letters with a partial or complete spelling. This provides more attestations of the syllabic spelling of the divine name than any other attested source of the second millennium BCE. Of these, seventy-seven vocalize the deity’s name as a-ši-ir-te/ti/ta, while eighteen vocalize the name as aš-ra-tu/ti/ta, with the final vowel functioning as a case vowel.9 The choice of the name, abdi-a-ši-ir-te, is based on how the name bearer himself (or his scribe) writes his own name in the letters he is responsible for (EA 61 line 2; EA 62 line 2). However, even this can vary as he writes his name with the divine element as aš-ra-tu₄ in EA 60 line 2.

In thirteenth century BCE Ugarit, the divine name is attested more than sixty times, most frequently in the alphabet cuneiform mythological texts, and also in some ritual texts. There it is always spelled as ʾtrt. The honorific rbt, “Lady”, can precede it, and the epithet ym most often follows it. This is usually translated as “Lady ʾtrt of the Sea”.10

The name occurs in a personal name at Ugarit in the syllabic cuneiform and referencing the same name as the ruler of Amurru in the Amarna letters. There, it is twice vocalized as ^I(ÌR) abdi-a-šar-ti, although only partially preserved in each case: ^I(ÌR) abdi-a-ša[r-ti] and ^I(ÌR) abdi-a-šar-t[i].11 It also occurs once in full (^I[ÌR] abdi-a-šar-ti) as a witness on a land sale document.12

The divine name also occurs on some polylingual god lists. At two places in Ugaritica V it may appear, where one occurrence preserves only the initial a-, but the second includes most of the name and can be restored as [aš(?)-r]a-tum or as [a?- ši?-r]a-tu₄.13 There is also a god list attestation with the syllabic reading ^daš-ra-tum.14 Huehnergard recognizes the apparent difficulty of his restoration of the [a?- ši?-r]a-tu₄ in light of his support of the other restoration. However, he argues that this is the Akkadian form of the goddess’ name, a point supported by the publication of an Ugaritic god list parallel to the syllabic cuneiform list.15 In line 19, where the syllabic god list has ^daš-ra-tum, we find that the Ugaritic god list has [ʾat]rt. This parallels the syllabic reading ^daš-ra-tum in the already cited syllabic cuneiform god list. In the above mentioned land sale document the name bearer appears as the son of ia-an-ḫa-am-mi, an undoubted West Semitic personal name.16 The attempt to read šar (the SAR sign) as šir₉, because it is more common in W. Semitic peripheral Akkadian, is possible but not substantiated.17 Thus, the Old Babylonian ^da-ši-ra-tum preserves the original form of the divine name.18 Quoting Wilfred G. Lambert, Wiggins notes that this is not a deity of Akkadian origin because “Akkadian does not tolerate 3 short vowels separated by single consonants”.19 West Semitic syllabic orthography did not retain this form with an initial vowel and two internal vowels. Instead, it generated by forms already occurring in Old Babylonian as aš-ra-tum, and appearing as both aš-ra-tu/ti/ta and a-ši-ir-te/ta to identify the name of the fourteenth century B.C. leader of Amurru.

This seems likely based on syncope in Akkadian (and West Semitic) orthography. It occurs when the last short vowel, where there are two or more non-final short vowels in a series of open syllables, tends to drop.20 Thus, *parisum becomes parsum. Here, ^da-ši-ra-tum becomes ^da-šir-tum (where -tum is also read as -tu₄ with the loss of mimation after the Old Babylonian period) or a-ši-ir-te. Also possible is ^da-ši-ra-tum becoming ^daš-ra-tum. The latter is the less attested variant and, with Huehnergard, may be an Akkadian form.

Of special interest is the presence of the feminine suffix, either -tu or -atu, and the question of the final vowel and its operation as a case vowel. Although the final vowel on the abdi-a-ši-ir-te name varies, it does not function as a case vowel.21 Is the final -a the common ending for personal names from Amarna, regardless of case?22 In fact, at Amarna a generalization may be made for personal names with a West Semitic origin.23 Names in letters from Byblos, Amurru, Tyre, and Canaanite towns south of this area do function with case vowels. The exception is Jerusalem.24 The West Semitic personal names in these letters and those to the north of Byblos, Amurru, and Tyre do not have final vowels corresponding to the grammatical case in which they appear. Surprisingly, the alternation in final vowels attested in appearances of the name abdi-a-ši-ir-te in letters from Byblos, Amurru, and Tyre do not track with the grammatical case in which they appear. In contrast, a-zi-ri, the son of abdi-a-ši-ir-te, does appear with case vowels in the correspondence from Byblos and Tyre, but not in its twelve occurrences from Amurru.25 In the latter, it is always spelled a-zi-ri, including the correspondence from a-zi-ri himself. Thus, the operation of case vowels on some personal names in the Amarna texts is not only unpredictable, its presence or absence, as well as the presence of fixed forms, varies from one location to another. For abdi-a-ši-ir-te, this is true. The final syllable does, however, provide a feminine marker. Importantly, the most common final vowel is -a, occurring in 62 attestations, while 11 attestations have the final sign entirely missing and 22 have a final -e, -i, or -u vowel. Thus, nearly 74 percent of the known final vowels on the name abdi-a-ši-ir-te have the final -a vowel. Even if the other ten Bronze Age attestations are added, with one of the Ugarit Akkadian forms deducted because the final sign is entirely missing (resulting in nine occurrences), that still leaves 67 percent of all the Bronze Age names as ending in a final -a vowel. Thus, by a factor of two to one, this vowel dominates as the one used most often when writing the divine name. As noted, this cannot be explained as a case vowel, but is best understood as the common vocalic ending for this divine name, and perhaps for other names.26

Two important notes should be added regarding the West Semitic Bronze Age attestation of the name of the goddess. First, at Taanach a small archive of fourteen or fifteenth century BCE syllabic cuneiform tablets (TT 1–14) was discovered during the excavations of Ernst Sellin in 1903–1904. The cuneiform texts were published by F. Hrozný. TT 1 is a letter demanding fifty shekels of silver in payment.27 Hrozný read the first four signs on line 21 as ^da-ši-rat. This reading was accepted by later scholars.28 However, the unusual use of the -rat sign (elsewhere never used in the spelling of the goddess’ name) and the absence of a final vowel created a reading that has no parallel in the extant attestations of this deity in syllabic cuneiform. Basing his study on the notes of E. I. Gordon and Albert E. Glock, Anson F. Rainey proposed reading the line as ^GIŠza-ar-ni-nu, a type of wood and a Hurrian term, reflecting the Hurrian (or “northern”) influence in both the personal names and other Hurrian vocabulary in this letter.29 This proposal has been followed by others.30

The second note has to do with the publication of two Amorite–Akkadian bilinguals that appeared in January of 2023. If the first note dismisses a proposed reading of the goddess, this publication identified a clear syllabic cuneiform attestation.31 The two tablets originate in two separate private collections, one now in New York and the other in London. They were photographed using the Ammonium Chloride method. With a landscape format, the tablets date from the Old or Middle Babylonian period, i.e., the Middle or Late Bronze Ages. Using an Old Babylonian ductus encountered in tablets from Southern Babylonia during the period of Rīm-Sîn and of Hammurapi, these observations would push the date back to the first half of the second millennium BCE, perhaps the eighteenth century. Each tablet contains two columns in syllabic cuneiform. The first column (on the left side) appears to be in a language related to later West Semitic, likely a form of Amorite. The second column is Old Babylonian. The first ten lines of the first tablet contain a list of ten largely West Semitic deities. After identifying Dagan and Kamish in lines one and two, the third line names the first female deity, read as a-še-˹ra˺-tum.32 The corresponding Akkadian divine name in the right column is written DIĜIR.MAḪ, which may be transliterated into Akkadian as Bēlet-ilī “Lady of the gods”. This is the Mesopotamian mother-goddess. As is generally accepted, authors George and Krebernik read the internal vowels of a-še-˹ra˺-tum as short.

The point noted here is that the spelling, ašeratum, corresponds to the form already proposed as the best way to explain the development into the other attested spellings, ašratum and ašertum. In this list, the most common vocalic ending for the Amorite deities is -u(m) (lines 3–8), possibly reflecting the nominative case ending. The same is true in all the remaining single nouns found in lines 13–21 (after which phrases are listed on each line). As noted, however, by far the most common vocalic ending for the divine name Asherah/Asherata is -a. Therefore, it may be proposed that, behind the variants, there lay a basic form of this West Semitic divine name in the Bronze Age, that is, Ašerata. This name is thus clearly attested in the West Semitic of the Bronze Age. It will be argued below that this vocalization is the best way in which to understand the Northwest Semitic texts of the first millennium BCE, which, of course, preserve no vocalization because they are spelled in alphabetic script.33

2. Iron Age Evidence

The term ʾšr(h/t) appears in Iron Age Northwest Semitic to refer to a cult site, a shrine, and a divine symbol.34 As a goddess, the name does not occur in Phoenician or the Northwest Semitic dialects surrounding Israel and Judah. Despite the abundance of Phoenician inscriptions, it appears neither as a divine name nor as a theophoric element in a personal name.35 The same is now true of the Ekron inscription, lʾšrt, which should be interpreted as a shrine.36 The only extra-biblical occurrences that may refer to a deity in the Iron Age are those found at Kuntillet ʿAjrud and Khirbet el-Qom. These are also questioned by scholars who see no evidence for a deity but rather a cult symbol that represented Yahweh.37 However, arguments in favor of a referent distinct from that of ʾšr(h/t), as found elsewhere, include the unique spelling of these occurrences and their contextual usage in an environment different from the others.

It is important first to collect the evidence. At Kuntillet ʿAjrud, the following occurrences of the goddess’ name appear, followed by the editors’ translation of the name and the larger context38:

3.1 line 2, wlʾšrth “I have blessed you to YHWH of Shômrôn (Samaria) and to His asherah”

3.6 line 7, wlʾšrth “I have blessed you to YHWH of Têmān and His asherah”

3.9 line 1, wlʾšrth “[…] to YHWH of Têmān and His asherah”

4.1.1 line 1, wlʾšrth “[…] recount to YHWH of Têmān and His asherah […]”

At Khirbet el-Qom, ʾšrth appears three times on a tomb inscription, where it is part of the following context39: (1) ʾryhw h?śr ktbh (2) brk ʾryhw lyhwh (3) wmṣryh lʾšrth hwšʿ lh (4) ldnyhw (5) wlʾšrth (6) [wlʾš]r[t]h.

Aḥituv translates “Uriyahu the governor wrote it: ‘Blessed by Uriyahu of YHWH <and> of His Asherah. And from his enemies grant him deliverance, to Daniyahu…and of his Asherah [and of] His [Ashe]rah.’”40

Alternatively, Ziony Zevit renders it “Uryahu, the prosperous, his inscription ([or] an inscription). I blessed Uryahu to Yahweh, to wit, from his enemies…for the sake of Asheratah save him …by Abiyahu…?? and to Asheratah…Asheratah”.41

What strikes the reader immediately is that most occurrences attested in these two sites follow a waw conjunctive, but all then begin with the inseparable preposition, lamed. Further, all follow the preposition with the same five consonants, a sequence of letters that is not found elsewhere in the West Semitic Iron Age inscriptional evidence. This includes the final -h, which raises the problem of its interpretation as a pronominal suffix, “his Asherah”. If so, Hebrew grammar would discourage understanding the noun as a divine name because personal names do not take a pronominal suffix.42 This is true for human individuals, but there may be a difference with deities whose identities could allow a deity to simultaneously co-exist in different places, distinguished by an additional qualifier, such as Yahweh of Samaria and Yahweh of Teman.43 As yhwh ṣ^ebāʾôt, this transcendent deity remains qualitatively different from local and regional manifestations. Thus, female deities may be linked to male deities (who are themselves qualified by specifications of geography, for example) with the use of “his”, as in the parallels at Ebla: “Rashap of Adani and his Adamma”.44 Nevertheless, Thomas, in his significant discussion of this point, may be correct in identifying the important role of the structures of both the Kuntillet ʿAjrud and the Khirbet el-Qom attestations as ultimately pointing toward a non-deity with a pronominal suffix for ʾšrth.45 However, one has the sense of nuanced distinctions that become increasingly unlikely to have been worked through in such detail by the composers of these graffiti-like texts. All this is moot if ʾšrth has no demonstrable pronominal suffix, as argued here.

In addition, the use of third masculine singular verbs in conjunction with the blessings at Kuntillet ʿAjrud suggest that the focus is the subject Yahweh and not ʾšrth.46 It has been assumed that ʾšrth is not the name of a goddess but must be something else, a cult object or shrine.47 However, many retain the view that this is the deity Asherah.48 They assume that in some manner the grammar is not an insuperable barrier, often citing examples of exceptions from Ugarit or biblical texts.49 The structure “brk + l + noun” always identifies the noun as a divine name of a god or goddess.50 Further, Yahweh may retain the main focus of the verbs of blessing without excluding the presence of a consort.51

In fact, lʾšrth could preserve a final vowel letter, either an -a or an -o, that has no connection with a pronoun. Former studies considered the -tâ suffix to be an example of a double feminine ending, comparing place names such as ʾeprātâ (Gen. 35:16, 19; 48:7; Mic. 5:1; Ps. 132:6; Ruth 4:11; or as a personal name 1 Chr. 2:19, 24, 50; 4:4).52 However, the existing parallels in Hebrew can either be understood as accusatives of direction or as final -â vowels serving as an adverbial/accusative ending.53

We have already noted from the Late Bronze Age attestations the manner in which this Asherata form normally ends, in an -a vowel, following a -t-. It should therefore be no surprise to discover that the final vowel here is an -a.54 The ʾšrth is situated in the same tradition as the Asherata attestations at Amarna, Ugarit, and earlier.

The final -â has been explained by others as an archaic diptotic ending for the genitive and accusative cases.55 Examples of diptotic case endings may be found in the Amurru correspondence and at Ugarit, where a-na ^Irap-a-na occurs, in which the a-na corresponds to the Hebrew l- preposition as used at Kuntillet ʿAjrud and Khirbet el-Qom.56 Hebrew orthography can indicate 8th and 7th century BCE identification of a final -a by means of a final -h.57

The ending on ʾšrth may provide an example of the final –(t)a vocalization as a feminine suffix marker. The place name examples of Israelite and Judean feminine names with this suffix may be recognized in at least two feminine geographic names that were identified on the list of names from the region that was conquered by pharaoh Shishak c. 925 BCE. These names are [P]nu=ʾI=ru Ḥa=d=ša=ta “New Penuel” or ʿA=ru=d=ʾa=t Ru=bi=ta “Great Arad”.58 The latter site lies roughly in the vicinity of Kuntillet ʿAjrud and Khirbet el-Qom. If the -ta served as a feminine marker in these two examples at the end of the tenth century, it may have served as a feminine ending more than a century later in the examples of ʾšrth. However, by the time of the Hebrew Bible’s language updating in the later eighth and the seventh centuries BCE, the feminine ending had dropped the -t. This phenomenon occurs in Old Aramaic, Samalian, and the Deir Alla texts.59 It also appears in the Egyptian transcription of at least ten Iron Age Hebrew words, in comparison with their biblical Hebrew counterparts: ʾtpāta (Heb. ʾašpâ “quiver”); ʿalīta (Heb. ʿālîyâ “upper chamber”); ʿagálta (Heb. ʿāgālâ “wagon”); ʿaduta (Heb. ʿēdâ “conspiracy”); baʾlat (Heb. baʿālâ “lady”); markábata (Heb. merkābâ “chariot”) maģārata (Heb. mĕʿārâ “cave”); ḫilbatta (Heb. ḥelbĕnâ “perfumed water”); śaʿ(ĕ)rata (Heb. śaʿărâ “hair; wool”); and qarta (Heb. qîryâ “town, city”).60

The final -t as an archaic feminine marker is preserved in feminine names in BH: bośmat, ṭopat, moḥălat, gînat, and šimʿāt.61 The first three personal names occur either in Genesis (chs. 26, 28, and 36) or in the tenth century Solomonic period and the generation after Solomon. The final two names, which appear later (ninth and early eighth centuries BCE), are not certainly feminine. moḥălat also appears as maḥlâ, a daughter of Zelophehad, in Num. 26:11, 33; 27:1; Josh. 17:3. Thus, the feminine suffix marker appears as both -t and -â in a Hebrew name that is otherwise spelled in an identical fashion.62

McClellan suggests that Tropper’s observation concerning shortened forms of YHWH with final -a on the ends of theophoric personal names can be transferred to the divine name ʾšrth.63 However, the latter name does not appear in the Iron Age as a theophoric element forming part of personal names. Instead, it is independent.

The final –(t)a vowel has already been observed at the end of Bronze Age appearances of the divine name, Asherata. Indeed, a review of all the evidence is consistent with the understanding that the fully vocalized name of the goddess, Asherata, now attested in the Old Babylonian Amorite bilinguals, is preserved in the alphabetic form ʾšrth, which is found in all the Iron Age IIB Hebrew occurrences of the divine name that are attested outside the biblical text.

The spelling of ʾšrh in the Biblical attestations, whether of a (wooden?) symbol of the goddess (e.g., Jdg. 6:28; 2 Kgs. 13:6) or of the divine name itself (e.g., 1 Kgs. 18:19),64 seems to represent an updating of the Hebrew text in the late Iron Age (late eighth, seventh, and early sixth centuries BCE).65 By this time, at least in the Jerusalem dialect of Hebrew, which is represented by much of the Hebrew Bible in its latest pre-exilic updating, the -t had lost its symbol as a suffix marker of the feminine, while the final -â (marked in the alphabetic script by -h) represented this phenomenon.66

Because the Kuntillet ʿAjrud and Khirbet el-Qom inscriptions predate this change, lʾšrth should be understood as in agreement with the unified witness of the Bronze Age evidence for the name of the goddess, as well as the Iron Age inscriptional evidence of Shishak’s proper names in and around Southern Canaan and the evidence of multiple feminine cognate Hebrew nouns preserved in Egyptian that all read a final -ta as either a preserved form for Asherata or as an example of other names and nouns of the same construction.

In conclusion, both the addition of the Amorite bilingual texts at the beginning of the appearance of the name Asherata and the full publication of the Kuntillet ʿAjrud texts in the Iron Age period that attest to all the ʾšrth names there affirm that the goddess known as Asherah in the Bible was consistently and best rendered as Asherata. There is no third person singular suffix attached to this divine name, no “his” here. The divine names do not take the pronominal suffixes at Kuntillet ʿAjrud and Khirbet el-Qom. These attestations are best understood as the goddess, known in the Bible as Asherah, and secondarily as a cult object.