Diglossia in Ancient Hebrew

Abstract

The present article summarizes the research assembled in my book Diglossia in Ancient Hebrew (1990), which in turn was based on my doctoral dissertation of a decade earlier, “Evidence for Spoken Hebrew in Biblical Times” (1980). The following linguistic issues are surveyed: gender neutralization (in both verbs and pronouns), the merger of III-ʾ and III-y verbs, the adjectival clause, and the genitive exponent.

In Memoriam, Avi Hurvitz (1936–2025), יהי זכרו ברוך.

1. Introduction1

Thirty-five years have passed since the publication of my book, Diglossia in Ancient Hebrew (Rendsburg 1990), which in turn was based on my doctoral dissertation of a decade earlier, “Evidence for Spoken Hebrew in Biblical Times” (Rendsburg 1980). The thesis of the book and its predecessor dissertation was relatively simple: Standard Biblical Hebrew (SBH) and even Late Biblical Hebrew (LBH), the language in which the biblical books were composed, represent the standard literary dialect of ancient Hebrew through the ages; whereas antecedents of Mishnaic Hebrew (MH) and other divergences from the grammatical norm in the Bible represent the spoken dialect of ancient Hebrew. As such, ancient Hebrew reflects diglossia, in its original sense as propounded by Charles Ferguson (1959, p. 336):

Diglossia is a relatively stable language situation in which, in addition to the primary dialects of the language (which may include a standard or regional standards), there is a very divergent, highly codified (often grammatically more complex) superposed variety, the vehicle of a large and respected body of written literature, either of an earlier period or in another speech community, which is learned largely by formal education and is used for most written and formal spoken purposes but is not used by any sector of the community for ordinary conversation.

In sum, languages often display two registers, one high (H) for literary and formal purposes, and one low (L) for colloquial and informal purposes.2

Ferguson presented four examples from amongst the world’s languages: spoken Arabic vs. classical Arabic, Swiss German vs. standard German, Haitian Creole vs. standard French, and spoken Greek vs. literary Greek. Moreover, he wrote, “It is likely that this particular situation in speech communities is very widespread, although it is rarely mentioned, let alone satisfactorily described” (Ferguson 1959, p. 326). Roy Andrew Miller, within the context of discussing diglossia in Japanese, went even further, “probably no language on earth has ever been written as it is spoken” (Miller 1967, p. 138).3

As adumbrated above, to my mind, the standard Hebrew (SBH to be sure, though LBH as well) used to compose the biblical books (replete with wayyiqṭol forms and so on) represents the literary dialect of ancient Hebrew—in fact, as Chaim Rabin (1979) proposed, perhaps even a literary artifice created by Israelite bookmen and literati at work in the Davidic-Solomonic chancellery.4 The colloquial dialect would have been much closer to what emerges centuries or even a millennium later as MH, but whose seeds may be found within the biblical text, sometimes as stylistic usage (perhaps), though most often simply as incursions from the spoken patois into the literary text. Strikingly, as the following will demonstrate, many of these colloquial usages find parallels in spoken Arabic (and at times in other modern spoken Semitic languages).5

My work was not the first to propose diglossia in ancient Hebrew, for other scholars already had proposed or theorized such, as follows. Edward Ullendorff (1971, p. 249) wrote:

Perhaps we ought to go a good deal further than that and recognize that the language of the Mishna, principally derived from those oral sources, was in fact the product of the colloquial used during the Biblical period, the amalgam of its standard and sub-standard versions. Perhaps Mishnaic is the vernacular so rarely encountered in the predominantly (though by no means exclusively) formal language of the OT [Old Testament]? Perhaps BH [Biblical Hebrew], in its Masoretic garb, is simply the literary counterpart to the Mishnaic colloquial? This assumption of the contrasting, but not necessarily temporally divided, functions of Mishnaic and BH would also explain the essentially prosaic character of Mishnaic, the absence of poetic words and forms, and the somewhat simplified morphology which has been greatly affected by the workings of analogy.6

The same great scholar also wrote, “Finally, some serious work ought to be undertaken, in the light of modern linguistic notions, on the question of dialects and colloquialisms in the Old Testament” (Ullendorff 1962, p. 465).7 Indeed, my dissertation and subsequent book were a direct response to Ullendorff’s recommendation.

Briefly I also may quote Stephen Lieberman (1978, p. 26), “The vernacular also left its traces in numerous slips from the literary norm in the written compositions of the period”; and William Chomsky (1964, p. 161), “More probably, men like Amos and Isaiah, after writing down or delivering their lofty and noble messages in the classical style, addressed their acquaintances or members of their family in the simple conversational dialect, including colloquialisms and slang, current among the rest of the population.” In addition, one must mention the work of Abba Bendavid (1951, 1967–1971), who did so much to highlight and categorize the differences between BH and MH, with the sometimes explicit though always implicit understanding that the latter represents the spoken variety reaching back centuries if not a millennium.8

Notwithstanding the above comments, as implied by Ullendorff’s call that “some serious work ought to be undertaken”, prior to my own research no one had truly identified specific colloquial features in the biblical text, certainly not in a systematic fashion.9 In what follows I present a summary of the material, or at least the most salient features.10

2. Gender Neutralization (Pronouns)

As is well known, Semitic languages distinguish gender in the 2nd and 3rd person plural pronouns. Such occurs regularly in BH, with m-forms used for the masculine and n-forms used for the feminine. And yet relatively frequently in the Bible one finds m-forms employed for the feminine, thus exhibiting gender neutralization, with no distinction between masculine and feminine forms. Since this same phenomenon occurs in spoken Arabic (see below), I conclude that these instances of gender neutralization in the Bible are features of ancient spoken Hebrew which have permeated into the written text. Examples follow:11

2.1. 2nd Person Pronominal Suffix

2 Sam 1:24 בְּנוֹת֙ יִשְׂרָאֵ֔ל אֶל־שָׁא֖וּל בְּכֶ֑ינָה הַמַּלְבִּֽשְׁכֶ֤ם שָׁנִי֙ עִם־עֲדָנִ֔ים

‘daughters of Israel, weep over Saul, who clothed you in crimson with fineries’

Amos 4:1–2 שִׁמְע֞וּ הַדָּבָ֣ר הַזֶּ֗ה פָּר֤וֹת הַבָּשָׁן֙ ... כִּ֛י הִנֵּ֥ה יָמִ֖ים בָּאִ֣ים עֲלֵיכֶ֑ם

       וְנִשָּׂ֤א אֶתְכֶם֙ בְּצִנּ֔וֹת

‘hear this word, O cows of Bashan … for days are coming upon you, and he will remove you with hooks’12

Song 2:7, 3:5, 5:8, 8:4 הִשְׁבַּ֙עְתִּי אֶתְכֶ֜ם בְּנ֤וֹת יְרוּשָׁ֙לִַם֙

‘I adjure you, O daughters of Jerusalem’13

2.2. 3rd Person Independent Pronoun

Jer 5:10 הָסִ֙ירוּ֙ נְטִ֣ישׁוֹתֶ֔יהָ כִּ֛י ל֥וֹא לַיהוָ֖ה הֵֽמָּה

‘remove her shoots, for they are not unto Yhwh’

Song 6:8 שִׁשִּׁ֥ים הֵ֙מָּה֙ מְלָכ֔וֹת

‘they are sixty queens’

Neh 3:34 הַיְחַיּ֧וּ אֶת־הָאֲבָנִ֛ים מֵעֲרֵמ֥וֹת הֶעָפָ֖ר וְהֵ֥מָּה שְׂרוּפֽוֹת

‘will they revive the stones from the rubbish heaps, even as they are burnt?’

2.3. 3rd Person Pronominal Suffix

Gen 26:15 וְכָל־הַבְּאֵרֹ֗ת אֲשֶׁ֤ר חָֽפְרוּ֙ עַבְדֵ֣י אָבִ֔יו ... סִתְּמ֣וּם פְּלִשְׁתִּ֔ים וַיְמַלְא֖וּם עָפָֽר

‘and all the wells which the servants of his father had dug … the Philistines blocked them and filled them with dirt’ (see also v. 18)

Gen 41:23 וְהִנֵּה֙ שֶׁ֣בַע שִׁבֳּלִ֔ים צְנֻמ֥וֹת דַּקּ֖וֹת שְׁדֻפ֣וֹת קָדִ֑ים צֹמְח֖וֹת אַחֲרֵיהֶֽם

‘and behold seven ears, shriveled, thin, scorched by the east-wind, are sprouting after them’ (see also v. 27 below in §3.2)

Exod 2:16–17   וּלְכֹהֵ֥ן מִדְיָ֖ן שֶׁ֣בַע בָּנ֑וֹת ... וַיַּ֖שְׁקְ אֶת־צֹאנָֽם

‘and the priest of Midian had seven daughters … and he (sc. Moses) watered their flock’

Num 27:7 כֵּ֗ן בְּנ֣וֹת צְלָפְחָד֘ דֹּבְרֹת֒ נָתֹ֙ן תִּתֵּ֤ן לָהֶם֙ אֲחֻזַּ֣ת נַחֲלָ֔ה בְּת֖וֹךְ אֲחֵ֣י אֲבִיהֶ֑ם

‘the daughters of Zelophehad speak correctly, indeed give to them an inheritance plot amongst the brethren of their father’14

1 Sam 9:20 וְלָאֲתֹנ֞וֹת הָאֹבְד֣וֹת לְךָ֗ הַיּוֹם֙ שְׁלֹ֣שֶׁת הַיָּמִ֔ים אַל־תָּ֧שֶׂם אֶֽת־לִבְּךָ֛ לָהֶ֖ם

     כִּ֣י נִמְצָ֑אוּ

‘and as for your lost jennies, for three days now, do not concern yourself (lit. ‘do not place your heart’) about them, for they have been found’

Jer 44:2 אַתֶּ֣ם רְאִיתֶ֗ם אֵ֤ת כָּל־הָֽרָעָה֙ אֲשֶׁ֤ר הֵבֵ֙אתִי֙ עַל־יְר֣וּשָׁלִַ֔ם וְעַ֖ל כָּל־עָרֵ֣י יְהוּדָ֑ה

     וְהִנָּ֤ם חָרְבָּה֙ הַיּ֣וֹם הַזֶּ֔ה וְאֵ֥ין בָּהֶ֖ם יוֹשֵֽׁב׃

‘you have seen all the evil that I have brought upon Jerusalem and all the cities of Judah; and behold they are a ruin today, and no one dwells in them’

Job 42:15 וְלֹ֙א נִמְצָ֜א נָשִׁ֥ים יָפ֛וֹת כִּבְנ֥וֹת אִיּ֖וֹב בְּכָל־הָאָ֑רֶץ וַיִּתֵּ֙ן לָהֶ֧ם אֲבִיהֶ֛ם נַחֲלָ֖ה

     בְּת֥וֹךְ אֲחֵיהֶֽם

‘and there were not found women as beautiful as the daughters of Job in all the land; and their father gave to them an inheritance amongst their brothers’

2.4. Arabic Evidence

As indicated above, Arabic exhibits the same pattern as Hebrew. In the classical language, gender distinction is retained throughout: 2nd. masc. pl. ʾantum, 2nd. fem. pl. ʾantunna, 3rd masc. pl. hum, 3rd fem. pl. hunna (independent pronouns); 2nd. masc. pl. -kum, 2nd. fem. pl. -kunna, 3rd masc. pl. -hum, 3rd fem. pl. -hunna (pronominal suffixes) (Fischer 1965, pp. 124–25; 2002, pp. 141–42). In most spoken dialects,15 however, m-forms are used irrespective of gender, e.g., Iraqi 2nd com. pl. ʾintu, 3rd com. pl. humma (independent pronouns), 2nd com. pl. -kum, 3rd com. pl. -hum (pronominal suffixes) (Erwin 1963, pp. 271–72).16 Note, moreover, that this feature is evident already in Middle Arabic, in both Jewish and Christian varieties (Blau 1966–1967, pp. 1.206–09; and more briefly 1980a, pp. 97–98; 1981, p. 90).

2.5. Ethiopian Evidence

As with all classical Semitic languages, Gəʿəz (classical Ethiopic) retains the gender distinction in the relevant pronouns: 2nd masc. pl. ʾantə́mu, 2nd fem. pl. ʾantə́n, 3rd masc. pl. ʾəmuntú, 3rd fem. pl. ʾəmāntú (independent pronouns); 2nd masc. pl. -kə́mu, 2nd fem. pl. -kə́n, 3rd masc. pl. -hómu, 3rd fem. pl. -hón (pronominal suffixes) (Tropper and Hasselbach-Andee 2021, pp. 51–52). The general system is retained in Tigre, Tirginya, and Western Gurage, whereas in Amharic, Harari, Argobba, Gafat, and Eastern Gurage gender neutralization occurs.17 One hesitates to use the term ‘diglossia’ for the situation in Ethio-Semitic,18 though the evidence of the South Ethiopian vernacular languages especially demonstrates the general trend towards gender neutralization in varieties of spoken Semitic.

2.6. Mishnaic Hebrew Evidence

Most scholars, the present writer included, aver that MH is characterized by gender neutralization (see, e.g., Kutscher 1982, p. 123; Pérez Fernández 1999, p. 18),19 though the picture is complicated by the fact that both m-forms and n-forms occur, for both masculine and feminine, in the various textual witnesses. The interchange between m-forms and n-forms is due to two separate causes: (a) a general merger of the phonemes /m/ and /n/ in final position, as seen in the form אָדָן alongside אָדָם ‘man, person’ (e.g., M. Bava Batra 6:6 דֶּרֶךְ בְּנֵי אָדָן ‘the way of humankind’)20; and (b) the influence of Middle Aramaic, which uses n-forms as opposed to m-forms throughout.21 This will explain the apparent randomness of the pronoun forms, as well as Kutscher’s comment that the distinctive masculine and feminine BH forms “merged into one form הֵן, הֵם for both” (1982, p. 123).

This issue aside, the antecedents of the gender neutralization inherent in MH are to be found in the spoken Hebrew of biblical times, as seen by the ample instances of m-forms in place of expected n-forms (with the above passages as a mere sampling).

3. Gender Neutralization (Verbs)

As with the pronouns, so also with the prefix-conjugation (PC) 2nd and 3rd person plural verbs, as well as the imperative (impv.) 2nd plural forms. In BH the gender distinction is normal and expected, as follows: 2nd masc. pl. tiqṭəlū, 2nd fem. pl. tiqṭolnå̄, 3rd masc. pl. yiqṭəlū, 3rd fem. pl. tiqṭolnå̄ (PC)22/2nd masc. pl. qiṭlū, 2nd fem. pl. qəṭolnå̄ (impv.). Once again, however, relatively frequently in the Bible one finds the masculine forms employed with feminine subjects, thus creating gender neutralization.23 Since this same phenomenon occurs in spoken Arabic (see below), I conclude that these instances of gender neutralization in the Bible are further features of ancient spoken Hebrew which have permeated into the written text. Examples follow:24

3.1. 2nd Person Prefix-Conjugation (PC) Verbs

Joel 2:22 אַל־תִּֽירְאוּ֙ בַּהֲמ֣וֹת שָׂדַ֔י

‘fear not, beasts of the field’

Song 2:7, 3:5 הִשְׁבַּ֙עְתִּי אֶתְכֶ֜ם בְּנ֤וֹת יְרוּשָׁ֙לִַם֙ ... אִם־תָּעִ֧ירוּ׀ וְֽאִם־תְּעֽוֹרְר֛וּ

‘I adjure you, daughters of Jerusalem … do not rouse and do not arouse (love)’ (see also 5:8, 8:4)25

3.2. 3rd Person Prefix-Conjugation (PC) Verbs

Gen 41:27 וְשֶׁ֤בַע הַֽשִׁבֳּלִים֙ הָרֵק֔וֹת שְׁדֻפ֖וֹת הַקָּדִ֑ים יִהְי֕וּ שֶׁ֖בַע שְׁנֵ֥י רָעָֽב׃

‘and the seven empty ears, scorched by the east-wind, shall be seven years of famine’ (see also v. 23 above in §2.3)

Lev 26:33 וְעָרֵיכֶ֖ם יִהְי֥וּ חָרְבָּֽה

‘and your cities shall be in ruin’

Judg 21:21 אִם־יֵ֙צְא֥וּ בְנוֹת־שִׁילוֹ֘

‘if the daughters of Shiloh come-out’

1 Kgs 11:3 וַיַּטּ֥וּ נָשָׁ֖יו אֶת־לִבּֽוֹ

‘and his wives turned-away his heart’26

Isa 19:18 בַּיּ֣וֹם הַה֡וּא יִהְיוּ֩ חָמֵ֙שׁ עָרִ֜ים בְּאֶ֣רֶץ מִצְרַ֗יִם

‘on that day there shall be five cities in the land of Egypt’

Esth 1:20 וְכָל־הַנָּשִׁ֗ים יִתְּנ֤וּ יְקָר֙ לְבַעְלֵיהֶ֔ן

‘and all the women shall give honor to their husbands’

3.3. 2nd Plural Imperative Forms

Isa 23:1 הֵילִ֣ילוּ׀ אֳנִיּ֣וֹת תַּרְשִׁ֗ישׁ

‘howl, O ships of Tarshish’27

Amos 4:1 שִׁמְע֞וּ הַדָּבָ֣ר הַזֶּ֗ה פָּר֤וֹת הַבָּשָׁן֙

‘hear this word, O cows of Bashan’28

3.4. Arabic Evidence

As intimated above, Arabic exhibits the same pattern as Hebrew. In the classical language, gender distinction is retained in PC and imperative plural verbs: 2nd. masc. pl. taqtulū, 2nd. fem. pl. taqtulna, 3rd masc. pl. yaqtulū, 3rd fem. pl. yaqtulna (PC); 2nd. masc. pl. (ʾu)qtulū, 2nd. fem. pl. (ʾu)qtulna (impv.) (Fischer 1965, pp. 103–4; 2002, pp. 117, 120). In the spoken language, however, the gender distinction typically breaks down, with masculine forms used for both genders, thus, for example, the following Egyptian Arabic forms: 2nd com. pl. tiktibu, 3rd com. pl. yiktibu (PC), 2nd com. pl. ʾiktibu (impv.) (Mitchell [1956] 1978, pp. 34–35).29

3.5. Ethiopian Evidence

As with all classical Semitic languages, Gəʿəz (classical Ethiopic) retains the gender distinction in the relevant PC and imperative verbs: 2nd masc. pl. təqattə́lu, 2nd fem. pl. təqattə́lā, 3rd masc. pl. yəqattə́lu, 3rd fem. pl. yəqattə́lā (PC); 2nd masc. pl. qətálu, 2nd fem. pl. qətálā (impv.) (Tropper and Hasselbach-Andee 2021, pp. 112, 114). The general system is retained in Tigre, Tirginya, and Western Gurage, whereas in Amharic, Harari, Argobba, Gafat, and Eastern Gurage gender neutralization occurs.30 Again, one hesitates to use the term ‘diglossia’ for the situation in Ethio-Semitic, though the evidence of the South Ethiopian vernacular languages demonstrates the general trend towards gender neutralization in varieties of spoken Semitic.

3.6. Mishnaic Hebrew Evidence

As with the pronoun, even more clearly with the PC and imperative verbs: the BH fem. pl. tiqṭolnå̄ (PC) and qətolnå̄ (impv.) forms disappear completely in MH, so that the formerly solely masc. pl. forms tiqṭəlū (PC 2nd masc. pl.), yiqṭəlū (PC 3rd masc. pl.), and qiṭlū (2nd masc. pl. impv.) serve for both genders (Kutscher 1982, p. 125; Pérez Fernández 1999, p. 106; Haneman 1979–1980, p. 35). A single example will suffice: M. Yevamot 4:10 וְכֵן שְׁאַר כַּל הַנָּשִׁים לֹא יִנָּשְׂאוּ וְלֹא יִתְאָרְסוּ עַד שֶׁיִּהיוּ לָהֶן שְׁלוֹשָׁה חֳדָשִׁים ‘and thus with all other women, they may not be wed and they may not be betrothed until three months have been (i.e., passed) for them’.31

As seen above, the antecedents of the gender neutralization inherent in the MH 2nd and 3rd person plural verbs (both PC and impv.) are to be found in the spoken Hebrew of biblical times, as demonstrated by the ample instances of tiqṭəlū, yiqṭəlū, and qiṭlū forms serving as the predicate for feminine plural subjects (with the above passages as a mere sampling).

4. The Merger of III-ʾ and III-y Verbs

The two verb classes, III-ʾ (= Hebrew ל״א) and III-y (= Hebrew ל״י), produce distinct paradigms in classical Semitic languages, such as BH and Classical Arabic, with the former class adhering more or less to the strong verb paradigm and with the latter class producing its own weak-verb paradigm.32 In both MH and colloquial Arabic, however, the two verb classes largely fall together, with the III-ʾ verbs conjugated according to the III-y pattern (see details below). Once again, one may find anticipations of this merger in the biblical text, as the following examples illustrate.33

4.1. Suffix-Conjugation (SC) Verbs34

1 Sam 6:10 וְאֶת־בְּנֵיהֶ֖ם כָּל֥וּ בַבָּֽיִת ‘and their calves (lit. ‘sons’) they shut-up in the house’—from the root כ-ל-א

1 Sam 10:6 וְהִתְנַבִּ֖יתָ עִמָּ֑ם ‘and you will speak-in-ecstasy with them’—from the root נ-ב-א

2 Kgs 2:21 רִפִּ֙אתִי֙ לַמַּ֣יִם הָאֵ֔לֶּה ‘I heal these waters’—from the root ר-פ-א (with the ʾaleph still written here, though note the vocalization according to the III-y pattern)

4.2. Prefix-Conjugation (PC) Verbs35

Gen 23:6 לֹֽא־יִכְלֶ֥ה מִמְּךָ֖ מִקְּבֹ֥ר מֵתֶֽךָ ‘(no-one) will prevent you from burying your dead’—from the root כ-ל-א

2 Kgs 2:22 וַיֵּרָפ֣וּ הַמַּ֔יִם עַ֖ד הַיּ֣וֹם הַזֶּ֑ה ‘and the waters were healed unto this day’—from the root ר-פ-א

Jer 8:11 וַיְרַפּ֞וּ אֶת־שֶׁ֤בֶר בַּת־עַמִּי֙ עַל־נְקַלָּ֔ה ‘they heal the wound of my people blithely’—from the root ר-פ-א

Jer 10:5 כְּתֹ֙מֶר מִקְשָׁ֥ה הֵ֙מָּה֙ וְלֹ֣א יְדַבֵּ֔רוּ נָשׂ֥וֹא יִנָּשׂ֖וּא כִּ֣י לֹ֣א יִצְעָ֑דוּ ‘they are like a scarecrow in a cucumber-patch, they do not speak, indeed they must be carried, for they do not walk’—from the root נ-שׂ-א (with the odd orthography including the final ʾaleph after the waw, though with the vocalization clearly matching the III-y pattern)

4.3. Imperative Forms

Josh 24:14 וְעַתָּ֞ה יְר֧אוּ אֶת־יְהוָ֛ה ‘and now, fear Yhwh’—from the root י-ר-א

1 Sam 12:24 אַ֣ךְ׀ יְר֣אוּ אֶת־יְהוָ֗ה ‘only fear Yhwh’—from the root י-ר-א

Ps 34:10 יְר֣אוּ אֶת־יְהוָ֣ה ‘fear Yhwh’—from the root י-ר-א

These three passages all follow the same wording, perhaps the commonness of the expression led to the ‘weakening’ (for lack of a better term) of the expected III-ʾ pattern, with the resultant merger with the III-y conjugation.

4.4. Participle Forms

1 Sam 14:33 הִנֵּ֥ה הָעָ֛ם חֹטִ֥אים לַֽיהוָ֖ה ‘behold, the people are sinning against Yhwh’—from the root ח-ט-א, with the ʾaleph retained in the orthography, though with the III-y conjugation present

Ps 99:6 קֹרִ֥אים אֶל־יְ֜הוָ֗ה ‘calling unto Yhwh’—from the root ק-ר-א, with the ʾaleph retained in the orthography, though with the III-y conjugation present

Qoh 10:5 כִּשְׁגָגָ֕ה שֶׁיֹּצָ֖א מִלִּפְנֵ֥י הַשַּׁלִּֽיט ‘like an error which goes-forth from the ruler’—from the root י-צ-א, with the ʾaleph retained in the orthography, though with the III-y conjugation present (in place of expected יֹצֵאת [Gen 24:15, 43, 45, etc.])

4.5. Infinitive Forms

Judg 8:1 לְבִלְתִּי֙ קְרֹ֣אות לָ֔נוּ ‘without calling us’—from the root ק-ר-א, with the ʾaleph retained in the orthography, though with the III-y conjugation present

1 Sam 10:13 וַיְכַל֙ מֵֽהִתְנַבּ֔וֹת וַיָּבֹ֖א הַבָּמָֽה ‘and he finished to-speak-in-ecstasy, and he came to the high-place’—from the root נ-ב-א

Ezek 5:2 כִּמְלֹ֖את יְמֵ֣י הַמָּצ֑וֹר ‘when the days of the siege are completed’—from the root מ-ל-א, with the ʾaleph retained in the orthography, though with the III-y conjugation present

Note that there are many such cases with the root מ-ל-א ‘fill, complete’, perhaps due to the sonorous quality of /l/ immediately before the glottal stop. In fact, the Qal form מְלֹאת occurs 13x, with no occurrences of the paradigmatic III-ʾ form; in the Piʿel, one finds five attestations of the III-y vocalization vs. six attestations of the expected III-ʾ vocalization.

4.6. Arabic Evidence

As observed above already, whereas classical Arabic retains the distinction between III-ʾ and III-y verbs, in the colloquial varieties the two verb classes generally are merged, with the former conjugated as the latter—thus, for example, the following two common verbs in Palestinian Arabic, r-m-y ‘throw’ and q-r-ʾ ‘read’ (showing just the singular SC forms):36

	r-m-y (III-y)	q-r-ʾ (III-ʾ)
1st com. sg.	ramēt	qarēt
2nd masc. sg.	ramēt	qarēt
2nd fem. sg.	ramēti	qarēti
3rd masc. sg.	rama	qara
3rd fem. sg.	ramat	qarat

The merger of the two verb classes occurs even in more ‘conservative’ dialects, such as Omani Arabic: “Verbs with final hamza lose the glottal stop and behave like final-glide verbs, e.g., qarā/yiqra ‘read’” (Morano 2022, p. 131). Once again, this phenomenon may be observed already in Middle Arabic, amongst all three religious groups, e.g., לם אקר ‘he did not read’ (Jewish: Blau 1980a, p. 84), اقرى ‘read!’, اخطينا ‘we have sinned’ (Christian: Blau 1966–1967, pp. 1.176–78), and اخطيكم ‘they cause you to sin’ (Muslim: Blau 1981, p. 125).37

4.7. Mishnaic Hebrew Evidence

As intimated above, the merger of III-ʾ and III-y verbs is a distinguishing feature of MH. To be sure, each manuscript needs to be judged on its own, and in fact only Parma A and the Cairo Geniza documents have received detailed analysis (see, respectively, Haneman 1979–1980, pp. 391–431; and Birnbaum 2008, per the index on p. 367). The surveys by these two scholars suggest that the merger occurs about half the time, while the distinction is retained about half the time.

Most scholars opine that the merger is essentially complete, as seen by the following statements: “The ל״א verbs usually become ל״י (Kutscher 1982, p. 127); and “In RH [Rabbinic Hebrew], lamed-alef verbs tend to be inflected as though they were lamed-he” (Pérez Fernández 1999, p. 114). Presumably, such scholars would explain the forms which retain the III-ʾ paradigm as historical vestiges, even as “authentic” MH reflects the more or less complete merger of the two verb classes. This is the approach that I use, pace Haneman (1979–1980, pp. 422–24), who argued that there is no need to doubt the authenticity of those III-ʾ forms which retain the standard paradigm. Regardless of how this point is decided, all scholars would agree that MH displays the merger of the two verb classes far more commonly than any other Hebrew register.

To provide the reader with a sense of such forms, herewith a few samples, culled from vocalized (or partially vocalized) Cairo Geniza documents (for which see Birnbaum 2008, with the relevant page numbers indicated):

• C.U.L. T-S E1.124—M. Kelim 28:6 (4x) מְטַמָּא (p. 100)
• C.U.L. T-S E1.124—M. Kelim 7:5 (2x) מְטַמִּין (p. 100)
• C.U.L. T-S E1.124—M. Kelim 7:4 (+2x) מְטַמִּאין (p. 100)
• Bodl. Heb. c. 23/2 (fols. 3–8)—M. ʿEruvin 2:1 יוֹצָא (p. 129)
• Bodl. Heb. d. 47/10 (fols. 23–30)—M. Kilʾayim 4:6 צָאוֹי (p. 162)
• Bodl. Heb. d. 47/10 (fols. 23–30)—M. Kilʾayim 7:6 נקרֶא (p. 162)
• NLI 5699(8) (olim Sassoon 523)—M. Maʿaśer Sheni 4:10 המוצֶא (p. 185)

Note how in almost all of these examples, the ʾaleph is still written, but the vocalization reflects the III-y pattern.

And then perhaps most famously, the opening line of the entire Mishna, namely, M. Berakhot 1:1 מֵאֵמָּתַיִ קוֹרִין אֶת שְׁמַע בַּעֲרָבִים ‘from when does one read the Shemaʿ in the evening?’, with קוֹרִין, derived from the root ק-ר-א though conjugated per the III-y pattern (for Parma A, see Haneman 1979–1980, p. 410).

4.8. Aramaic Influence?

Above I noted that in classical Semitic languages, the two verb classes are kept distinct, though there is one exception. In Aramaic, the distinction between III-ʾ and III-y verbs is attested during the Old Aramaic period (Degen 1969, p. 72), but from the Imperial Aramaic period onward the merger of the two verb classes is evident, with III-ʾ verbs assimilating to the pattern of III-y verbs.38

Due to this Aramaic development, some scholars have argued that the merger of the two classes in MH, along with its BH antecedents, is due to Aramaic influence (see, e.g., Wagner 1966, p. 138). Arguments against this position are as follows. First, many of the BH examples occur in earlier biblical books (see the passages in Genesis, Judges, and Samuel above), at the time when Aramaic still retained the distinction between the two verbal classes. Secondly, if Aramaic influence were at work, one would expect the merger to have occurred throughout post-biblical Hebrew, which is to say, not just in MH, but in Ben Sira and in Qumran Hebrew (QH) as well. But these literary compositions do not reveal the merger of III-ʾ and III-y verbs. True, without vocalization, it is difficult to know for sure, since, as we saw above, even in the Bible the ʾaleph may be retained in writing, even when the pronunciation reveals the shift to the III-y pattern.

But if we gauge the evidence strictly from the orthography (especially since we have no other method at our disposal), the point stands. To the best of my knowledge, Ben Sira has only one instance of a III-ʾ verb reflecting the III-y pattern, to wit, לרפאות ‘to heal’ in Ben Sira 3:28. As to QH, “to judge from the orthography, the two root types are almost always distinguished” (Reymond 2014, p. 189), with full supporting evidence, for example, “in the approximately fifty easily readable 3cp perfect verb forms that are from III-aleph roots in the nonbiblical scrolls … only one has a form reminiscent of a III-waw/yodh root (מלו ‘they fulfilled’ in 4Q401 22, 2)” (Reymond 2014, p. 189).

The only late literary corpus which shows a slight trend in the direction of the merger is the Samaritan Pentateuch (SP), for in Samaritan Hebrew (SH) certain III-ʾ verbs are conjugated as if they were III-y verbs (e.g., ח-ט-א). In general, however, the distinction of the two verb classes is retained, since SH III-ʾ verbs typically follow the overall pattern of III-guttural verbs (Ben-Ḥayyim 2000, pp. 131–32).

In sum, if Aramaic influence were at work, one would expect these post-biblical Hebrew corpora to reveal a similar array of III-ʾ verbs assimilating to the III-y pattern, but such is clearly not the case. Hence, one will assume that the merger of the two verb classes occurred in spoken Hebrew, as reflected in MH and with occasional antecedents in the Bible, though not in formal, written Hebrew.

5. The Adjectival Clause

In the classical Semitic languages which possess the definite article, the grammatical norm within the adjectival clause is for both noun and modifier to possess the definite article. Thus, for example, in BH Jon 1:2 הָעִיר הַגְּדוֹלָה ‘the big city’, and in Arabic الرجل الكريم ‘the noble man’ (Wright 1896–1898, p. 2.273). Such phrases may occur in the spoken dialects, but in both MH and colloquial Arabic the definite article is likely to be omitted from the noun.

5.1. Biblical Examples

We thus find in the Bible examples such as the following, with anarthrous substantive followed by determined adjective:

• 1 Sam 6:18 אָבֵל הַגְּדוֹלָה ‘the large meadow’
• 1 Sam 19:22 בּוֹר הַגָּדוֹל ‘the large cistern’
• 2 Sam 12:4 אִישׁ הֶעָשׁיר ‘the rich man’
• 2 Kgs 20:13 שֶׁמֶן הַטּוֹב ‘the good oil’
• Jer 6:16 דֶרֶךְ הַטּוֹב ‘the good way’
• Jer 6:20 קָנֶה הַטּוֹב ‘the good (i.e., sweet, fragrant) cane’
• Zech 4:7 הַר הַגָּדוֹל ‘the large mountain’

A specific usage of this type concerns fixed phrases related to particular places, such as gates and courtyards (all related to Jerusalem, though this may be expected given the centrality of the city to the Bible):

• 1 Kgs 7:8 חָצֵר הָאַחֶרֶת ‘the rear courtyard’
• 1 Kgs 7:12חָצֵר הַגְּדוֹלָה ‘the large courtyard’
• 2 Kgs 20:4 (Qeri) חָצֵר הַתִּיכֹנָה ‘the middle courtyard’
• Ezek 40:28 חָצֵר הַפְּנִימִי ‘the inner courtyard’
• Ezek 40:31 חָצֵר הַחִצוֹנָה ‘the outer courtyard’
• Ezek 9:2 שַׁעַר הָעֶלְיוֹן ‘the upper gate’

Per the approach taken in this article, I would explain these departures from the grammatical norm as colloquial usages (representative of L) which creeped into the standard literary composition (representative of H).

5.2. Arabic Evidence

As indicated above, the same grammatical development occurs in spoken Arabic dialects, per the following examples:

• bāb əš-šarʾi ‘the east gate’ (Grotzfeld 1965, p. 93) (Syria)
• yōm il-imbārak ‘the blessed day’ (Altoma 1969, p. 79) (Iraq)39
• sūq eṭ-ṭwīle ‘the long road’ (Feghali 1928, pp. 134–35) (Lebanon)
• ʿēn il-bēḍa ‘the white spring’ (Bauer 1913, pp. 105–6) (Palestine)
• dār əl-kbīra ‘the big house’ (Willms 1972, p. 55) (Mauritania)

Moreover, examples are to be found already in Middle Arabic, as follows:

• Judeo Arabic סכה̈ אלג̇דידה̈ ‘the new coin’ (Blau 1980a, p. 161; 1981, p. 107)
• Christian Arabic ارض القدسة ‘the holy land’ (Blau 1966–1967, pp. 2.359–60)

5.3. Mishnaic Hebrew Evidence

As further indicated above, MH attests to the same syntagma:40

• M. Kilʾayim 1:1 וּפוּל ^הַלָּבָן ‘and the white bean’41
• M. Kilʾayim 7:7 כְּפוּל ^הַלָּבָן ‘like the white bean’
• M. Ketubbot 8:7 פֵּירוֹת הַתְּלוּשִׁין מִן הַקַּרְקָע ‘the fruits separated from the land’42
• M. Ketubbot 8:7 פֵּירוֹת הַמְחוּבָּרִים לַקַּרְקָע ‘the fruits attached to the land’
• M. Bava Qama 1:2 רְשׁוּת הַמְּיוּחֶדֶת ‘the exclusive ownership’
• M. Ḥullin 4:7 נֶפֶשׁ הַיָּפָה ‘the good appetite’
• M. ʾAvot 1:1 אַנְשֵׁי כְנֶסֶת הַגִּדוֹלָה ‘the men of the great assembly’

As we saw in BH, certain set terms follow this pattern, e.g.,:

• M. Šeqalim 6:3 שַׁעַר הָעֶלְיוֹן ‘the upper gate’43

Finally, note the following string of terms, which may constitute a separate sub-category, in light of the gentilic adjectival forms, but which nonetheless presents the same syntactic pattern:

• M. Pesaḥim 3:1 כּוּתַּח הַבַּבְלִי שֵׁכַר הַמָּדִי חוֹמֶץ הָאֲדוֹמִי זֵיתוֹם הַמִּצְרִי ‘the Babylonian porridge, the Median beer, the Edomite vinegar, the Egyptian barley-beer’

In light of the usage in spoken Arabic, and given the consensus that MH constitutes a spoken variety of late antique Hebrew (with antecedents reaching back to the biblical period), one will conclude that the combination of anarthrous noun + determined adjective is a feature of spoken Semitic, both in ancient Hebrew and in colloquial Arabic.44 On the Hebrew side, accordingly, we have further evidence for diglossia in the ancient language.

6. Genitive Exponent

In the classical Semitic languages, the genitive (indicating possession and similar associations) typically is expressed via the affixing of a personal pronoun suffix to the noun (including to the nomen rectum of a construct phrase), or via the basic construct phrase, or via a circumlocution. Thus, for example, in BH one finds such basic usages as בֵּיתִי ‘my house’ (2 Sam 11:11), הַר־קָדְשִׁי ‘my holy mountain’ (lit. ‘the mountain of my holiness’) (Ps 2:6), עַבְדֵי־יִצְחָק ‘the servants of Isaac’ (Gen 26:19), הַצֹּאן֙ אֲשֶׁ֣ר לְאָבִ֔יהָ ‘the flock which was to her father’, i.e., ‘her father’s flock’ (Gen 29:9).45 In the spoken languages, however, frequently one encounters the independent genitive exponent, largely unknown in the classical Semitic languages. Examples include Hebrew שֶׁל, a standard feature of MH, and a series of forms which appear in the various Arabic dialects.46

6.1. Contrast Between BH and MH

The contrast between the two usages in BH and MH may be illustrated by comparing and contrasting what one may call the genitive of contents, as follows.47 In the former register, a construct chain occurs, for example:

• פַּךְ הַשֶּׁמֶן ‘the vial of oil’ (2 Kgs 9:1, 3)
• צַפַּחַת הַשֶּׁמֶן ‘the jar of oil’ (1 Kgs 17:14, 16)
• נֹאדוֹת הַיַּיִן ‘the skin-bottles of wine’ Josh 9:13 [see also v. 4])
• נֹאד יַיִן ‘a skin-bottle of wine’ (1 Sam 16:20)
• מִזְרְקֵי יַיִן ‘basins of wine’ (Amos 6:6)
• בְּאֵר מָיִם ‘a well of water’ (Gen 21:19)

In the latter, however, one finds the interposition of the form שֶׁל, thus for example:

• חָבִית שֶׁלַּיַּיִן ‘a jug of wine’ (M. Terumot 3:1, 10:3, etc.)
• חָבִית שֶׁלָּשֶּׁמֶן ‘a jug of oil’ (M. Terumot 8:10)
• חֲבִית שֶׁלַּ דְּבַשׁ ‘a jug of honey’ (M. Bava Qama 10:4; see also T. Bava Qama 10:28)48
• הבור של מים ‘the cistern of water’ (T. Terumot 8:9; see also 8:10)
• צלוחית של שמן ‘a dish of oil’ (T. Shabbat 3:5)
• כוס של יין ‘a cup of wine’ (T. Shabbat 3:7)
• ספל של מים ‘a mug of water’ (T. Shabbat 6:4, 13:4)

The usage of שֶׁל is so regular and ubiquitous in MH that Pérez Fernández was inspired to include it as among the “striking differences” between BH and MH, right at the start of his textbook, in his words, “genitive particle שֶׁל ‘of’, partially replacing the construct chain of classical Biblical Hebrew” (Pérez Fernández 1999, p. 1).49

Naturally the form שֶׁל occurs in other contexts, including with pronominal suffixes to express ‘my’, ‘your’ ‘his’, etc.50 Once again we may contrast BH and MH by assessing the following pair:

• Prov 7:23 כְּבֵדוֹ ‘his liver’ ~ M. Yoma 8:6 הַכָּבֵד שֶׁלּוֹ ‘its liver’

For another example, consider בְּתוֹכוֹ ‘in his/its midst’ (BH 10x) and בְּתוֹךְ שֶׁלּוֹ ‘in his allotment’ (MH 16x), even as the semantics are slightly different.

6.2. Arabic Evidence

As intimated above, the various Arabic colloquials also develop a genitive exponent, as exemplified by the following:51

• Algerian Arabic ntəʿ, e.g., rt̠əl ntəʿ əlḥəm ‘a pound of meat’ (Willms 1972, p. 38)
• Egyptian Arabic bitāʿ, e.g., il-kitāb bitāʿi ‘my book’, il-fingān bitāʿ il-ʾahwa ‘the cup of coffee’ (Abdel-Massih et al. 2009, pp. 15–16).
• Palestinian Arabic tabaʿ, e.g., el-bet tabaʿi ‘my house’ (Kapliwatzky 1968, p. 63)
• Iraqi Arabic māl, e.g., l-mudīr māl il-madrasa ‘the principal of the school’ (Altoma 1969, p. 83)
• Yemenite Arabic ḥagg, e.g., äl-bäyt ḥagg äl-ʿâmil ‘the house of ʿAmil’ (Rossi 1939, p. 19).

As these phrases illustrate, MH and spoken Arabic have undergone the same parallel development, with the use of the genitive exponent, a feature unattested in the classical language.

6.3. Ethiopian Evidence

Moreover, a similar development may be found in Tigre and Tigrinya (as noted in (Rendsburg 1990, pp. 121–22) and (Rubin 2005, pp. 53–54)), where the form nāy obtains, e.g., Tigrinya maṣḥāf nāy mamhər ‘the teacher’s book’ (lit. ‘the book of the teacher’).52 The comments about diglossia in Ethio-Semitic above are relevant here as well, though in the present instance we arrive nearer to that linguistic situation, given the closer relationship between the classical language Gəʿəz and the spoken North Ethiopian languages (notwithstanding the chronological issue). To be sure, the genitive exponent does not occur in the classical language, and yet it appears, as indicated, in the modern varieties spoken in the same geographical region.

6.4. Anticipations of the MH Usage in BH

The MH form שֶׁל did not emerge ex nihilo, rather its history may be traced back to the biblical period. During the Iron Age, one may assume that this particle was a feature of ancient spoken Hebrew, especially at home in northern Israel, which occasionally penetrated the biblical text composed in the classical, literary standard. Thus one finds:53

• 2 Kgs 6:11 מִ֥י מִשֶּׁלָּ֖נוּ אֶל־מֶ֥לֶךְ יִשְׂרָאֵֽל ‘who among us is to (sc. ‘on the side of’) the king of Israel’
• Jonah 1:7 בְּשֶׁלְּמִ֛י הָרָעָ֥ה הַזֹּ֖את לָ֑נוּ ‘on account of whom is this evil thing to us’
• Jonah 1:12 כִּ֣י בְשֶׁלִּ֔י הַסַּ֧עַר הַגָּד֛וֹל הַזֶּ֖ה עֲלֵיכֶֽם ‘for on my account is this great storm upon you’
• Song 1:6, 8:12 כַּרְמִ֥י שֶׁלִּ֖י ‘my vineyard’
• Song 3:7 מִטָּתוֹ֙ שֶׁלִּשְׁלֹמֹ֔ה ‘the litter of Solomon’
• Qoh 8:17 בְּ֠שֶׁל אֲשֶׁ֙ר יַעֲמֹ֧ל הָאָדָ֛ם ‘however much man may toil’

Various comments may be added to the basic presentation of the data, with the letter indicators referring to the six passages above:

(a)

While the form שֶׁל is clearly a Hebrew usage (and note that this passage occurs in an Israelian section of the book of Kings), quite possibly the specific form מִשֶּׁלָּנוּ has been placed purposefully in the mouth of the Aramean king.54 To be sure, the parallel usage זיל/דיל (including with pronominal suffixes) occurs in Aramaic, including in Sefire III 20 as זי לי ‘of mine’ (Fitzmyer 1995, pp. 140–41, 157), and then regularly in Egyptian Aramaic (Muraoka and Porten 1998, pp. 162–65).

(b and c)

Muraoka (2012) has made the keen observation that the first speech occurs in the mouth of the sailors when they speak to each other, and thus they use the form בְּשֶׁלְּמִי (v. 7), though when they approach Jonah they use a more elevated register with בַּאֲשֶׁ֛ר לְמִי־הָרָעָ֥ה הַזֹּ֖את לָ֑נוּ’ ‘on account of whom is this evil thing to us’ (v. 8). “Jonah, in his turn, apparently thought it more friendly and diplomatic to speak to the sailors in a form of Hebrew with which they would feel more at ease and at home” (Muraoka 2012, p. 131), and thus he responded to them using the form בְשֶׁלִּי (v. 12). All of this, avers Muraoka (and I agree), indicates diglossia, with vv. 7 and 12 displaying the vernacular usage and with v. 8 exhibiting the classic Hebrew idiom.

(d and e)

Most scholars assert that Song of Songs is a late book, and thus use the three-fold presence of שֶׁל in the poem to support that conclusion.55 While this remains possible, one should note that שֶׁל also has been deemed a feature of northern Hebrew, in addition to its development within the spoken language, and thus Noegel and Rendsburg (2009) conclude that Song of Songs is a product of the kingdom of Israel to be dated to the early monarchic period.56

(f)

Qohelet is the one clear book dated to the Persian period, by which point שֶׁל may have become more popular, as witnessed, for example, by its repeated use in 3Q15 Copper Scroll, composed centuries later, though prior to the florescence of MH.57

In sum, the genitive exponent שֶׁל was a feature of ancient spoken Hebrew, especially in northern Israel, though in time in southern Judah as well. It appears only occasionally in the biblical text, commonly in 3Q15 Copper Scroll, and then as a characteristic trait of MH. One will assume that ancient written Hebrew eschewed the form, as it retained the more classical usages described at the start of this section.

7. Conclusions

In my Diglossia book (Rendsburg 1990), I treated still other features of ancient spoken Hebrew, e.g., the relative pronoun -שֶׁ (related to the previous item), the anticipatory pronominal suffix, and the elision of he in Hiphʿil (and other) infinitives. Space considerations do not allow me to repeat all that information here, though the interested reader may wish to consult the monograph.58

The main conclusion of this study is that ancient Hebrew was characterized by diglossia. To repeat the words from the Introduction above: both SBH and even LBH, the language in which the biblical books were composed, represent the standard literary dialect of ancient Hebrew through the ages; whereas antecedents of MH and other divergences from the grammatical norm in the Bible represent the spoken dialect of ancient Hebrew.59

7.1. Ancient Spoken Hebrew and Other Spoken Semitic Varieties

In almost all cases—and certainly in the examples provided herein—these divergences from the grammatical norm are to be found in other varieties of spoken Semitic (especially colloquial Arabic, though at times also various spoken Ethiopian languages). Thus, gender neutralization in the pronoun (§2), gender neutralization in the verb (§3), the merger of III-ʾ and III-y verbs (§4), the adjectival clause, comprised of anarthrous noun + determined adjective (§5), and the genitive exponent (§6) are found not only in MH and in antecedents thereof within BH, but also in spoken Arabic (and in three of five instances also in various spoken Ethiopian languages).

We thus should ask how and why such diverse languages within the Semitic family reflect such similar usages within their spoken varieties. Here I would invoke the doctrine of parallel development, a linguistic phenomenon described by Antoine Meillet in his celebrated essay “Note sur une difficulté générale de la grammaire comparée” (Meillet 1921, pp. 36–43). His main thesis was that related languages, even after they have separated from the proto-language, incline to develop in the same direction, even without mutual contact. Meillet applied this finding to problems in comparative Indo-European grammar, and it then was introduced into Semitic studies by Joshua Blau in several of his publications (Blau 1969, 1980b).60

The spoken languages discussed herein, with large gaps of time and space between them, impel us to apply the rule of parallel development to explain their shared features. I conclude, therefore, that ancient spoken Hebrew and modern spoken Arabic (and where relevant the Ethio-Semitic languages as well) have independently developed similar characteristics, diverging respectively from their classical counterparts in similar manners (see Rendsburg 1991). As such, the linguistic features presented above are, I trust, sufficient evidence by which to recognize a certain drift in spoken Semitic, to be explained by the phenomenon of parallel development.

7.2. Statistical Analysis of Colloquial Features in the Biblical Text

Finally, we return to Hebrew per se, with some reflections on the biblical texts themselves. I concluded my Diglossia book (Rendsburg 1990) with a statistical survey of the features of ancient spoken Hebrew vis-à-vis the different biblical books and genres. Naturally, the statistics can reveal only general trends, since we are not dealing with an exact science—especially regarding the dates of biblical books and their genre classifications. That said, several trends do emerge, as follows.

First, spoken features are less likely to appear in poetry (including prophetic speeches) and in legal and cultic language, and are more likely to appear in narrative prose compositions. This finding probably could be predicted, since the former literary styles often are guided by formal styles and tight strictures, whereas the latter study is somewhat freer and may allow for more subconscious slips from the classical idiom (Rendsburg 1990, p. 157).

Second, within narrative prose, spoken features are as likely to occur in the voice of the narrator as they are to appear in quoted speech. Or to word this differently, colloquial features do not occur within quoted speech at a higher rate than they do in the voice of the narrator. If many of my examples above occur in quoted speech, this is only because biblical narrative prose is characterized by a remarkably high degree of quoted speech, in fact, c. 43% (as confirmed by a series of independent studies conducted by different scholars, myself included).61 To repeat, my analysis reveals that the prose authors couched the dialogues of the biblical characters not in colloquial Hebrew but rather in the standard idiom reserved for literary composition (Rendsburg 1990, p. 161).62 As such, this is analogous to the direct quotations in the Qurʾān, which appear in al-fuṣḥā (classical Arabic), for example, via the use of the dual, the nominal case system, etc.,63 not in al-ʿāmmiyya (colloquial Arabic).64

Third, with the passage of time more colloquial usages crept into the biblical text, so that SBH is less prone to such incursions, whereas LBH shows a greater propensity for their appearance. This trend may be explained by the loss of the official literary circle in Jerusalem after 586 b.c.e., all tied, of course, to the great social, political, and indeed religious upheavals in the wake of the historical events of the 6th century b.c.e. (Rendsburg 1990, pp. 167–68).