Edition 2.0

12 December 2011


To print use PDF file here


Biblical Hebrew Poetry and Word Play

Reconstructing the Original Oral, Aural and Visual Experience

By David Steinberg


Home page http://www.houseofdavid.ca/

Return to Table of Contents


VI Reconstruction of Pre-Exilic Biblical Hebrew (EBHP)

1. Aims in Reconstructing EBHP

Box 9 - Can Biblical Texts be Linguistically Dated?

2. Changes in the Pronunciation Tradition of Biblical Hebrew Between EBHP and that Recorded in the Tiberian Masoretic Tradition (early 10th century CE)

Box 10 - Justification of Proposals for EBHP

3. Guidelines I Have Used in Reconstructing EBHP

Table 10 - Mono-syllabic Prepositions and Conjunctions Usually Linked to the Following Word in the MT by a maqqeph/makef (מקף)

Table 11 - Vowels of EBHP

4. Examples of the EBHP Vocalization of Biblical Hebrew Texts

a. Archaic or Archaizing Poetic Texts

i)          Blessing of Jacob (Genesis 49:1-27)

ii)         Song of the Sea (Exodus 15:1b-18)

iii)        The Oracles of Balaam (poetic portions of Numbers 23:7-24:24)

iv)        Haʾazinu (Deuteronomy 32:1-43)

v)         Blessing of Moses (Deuteronomy 33)

vi)        Song of Deborah (Judges 5)

b. Various Short Poems: Genesis 2:23; Genesis 3:14-19; Genesis 4:6-7; Genesis 4:23b-24; Genesis 8:22; Genesis 9:6; Genesis 9:25-27; Genesis 12:2-3; Genesis 14:19-20; Genesis 16:10-12; Genesis 24:60; Genesis 25:23; Genesis 27:28-29; Genesis 27:39-40; Genesis 35:10-12; Genesis 48:15-16; Genesis 48:20; Exodus 32:18; Numbers 6:24-26; Numbers 10:35-36; Numbers 12:6b-8a; Numbers 21:14,15,17-18; Numbers 21:27-30; Joshua 10:12-13 (poetic portion); Judges 9:8-15; Judges 14:14, 18; Judges 15:16 (poetic portion); Judges 16:23-24 (poetic portion); 1 Samuel 15:22b-23; 1 Samuel 18:7 (poetic portion); 2 Samuel 3:33-34 (poetic portions); 2 Samuel 20:1 (poetic portion); 1 Kings 8:12-13; 1 Kings 12:16 (poetic portion); 2 Kings 19:21b-28; 2 Kings 19:31; 2 Kings 19:32b-34.

c. Psalmic Poetry

i) II Samuel Chapt. 22 (Second version Psalm 18) -

ii) Psalm 23

iii) Psalm 114

iv) Psalm 121

v) Psalm 122

vi) Psalm 130

d. Lamentations

i) Lament of David (II Samuel 1:19-27)

ii) Lamentations 3:1-15 ("Qinah meter")

e. Poetry of Song of Songs - Song 2:1-17

f. Poetry of Job - Job 3:3-10

g Prophetic Poetry

i) Jer. 1: 11-12; Jer. 1: 18-19; Jer. 19:14-15; Zeph. 3:1-2; Deut 15:1,4

ii) Amos 3:3-6; 3:8; 5:5-7; 5:10-12; 5:16b-17; 6:12; 8:7-10; 9:5-6; 9:13

h. Prose Texts

i) Genesis 2:18-24

ii) Genesis 4:1-3; Genesis 13:4-14; Joshua 7:1-3

iii) Siloam Inscription


VI Reconstruction of EBHP

1. Introduction

It goes without saying that the pronunciation of pre-exilic Biblical Hebrew (c. 1000-600 BCE) varied with "...socio-economic class, professional standing, degree and type of education, religious affiliation, ethnic origin, generation, and even sex."[1] We should aim at recovering, as closely as possible, the pronunciation that a scribe in Jerusalem 700-600 BCE would have used in reading poetry to upper class Judeans or members of the kings court ([EBHP]). For poems of northern origin this might have included some features of northern pronunciation which would share some of the phonetic features of Phoenician and Aramaic such as the contraction of diphthongs. The clearest example of such a poem is the Song of Deborah.

Scribes trained in Jerusalem 700-600 BCE were likely the authors of the bulk of surviving JEH e.g. Siloam Inscription, Lachish ostraca, Arad ostraca etc. The same circles were likely the composers and/or transmitters of most of the pre-exilic biblical texts. JEH documents have been preserved in their original language and orthography and, within limits, can serve as a guide to pronunciation. Except for archaisms used in poetry, the pre-exilic biblical texts would very likely have conformed to the norms of JEH.

I aim to do the following listed in rough order of importance:

(1) Distinguish the consonantal and vowel phonemes and indicate their likely pronunciation. This will require, among other things, differentiating between:

  long (geminated)[2] and short consonants;

  different qualities of vowels with emphasis on qualitative differences that are phonemic; and,

  between diphthongs, long vowels (phonological or phonetic[3]), short vowels and the absence of vowels.

(2) Establish the number of syllables and their boundaries and syllable length; and,

(3) Establish the syllable carrying the word stress (primary or secondary).


This will require an understanding of:

i) Pronunciation the main differences between:

  the probable phonology and use of vowel letters of Biblical Hebrew at time of writing;

  the pronunciation tradition embodied in the Tiberian vocalization; and,

  Hebrew as it is pronounced in modern Israel.

ii) Script and Orthography:

  the appearance of the text in different historical periods and the latitude this provided for mistakenly replacing one letter by another; and,

  the development of orthography and its impact on the range of meanings and pronunciations that could be attributed to the original consonantal skeleton.


Box 9

Can Biblical Texts be Linguistically Dated?[4]

After almost three centuries of modern study of the Hebrew Bible, it is clear that internal analysis of the text cannot convincingly disclose the periods of composition of the components that were finally redacted into the text that has come down to us. For this reason, the dating of textual units on objective linguistic grounds, if it can be shown to be feasible, would prove invaluable to the study of the Hebrew Bible and Israelite/Jewish history.

For decades, Biblical Hebrew texts have been roughly divided into three chronological strata based on linguistic criteria Archaic Biblical Hebrew (c. late second to early first millennia BCE), Classical or Early Biblical Hebrew (c. ninth to early sixth century BCE), and Late Biblical Hebrew (c. after the sixth century BCE). The most important research supporting this structure was done by Avi Hurvitz[5].

In the last decade this structure has been attacked by a number of scholars who maintain that the dating of Biblical Hebrew texts on the basis of language is effectively impossible. The most important books making this case are - Linguistic Dating of Biblical Texts by Ian Young, Robert Rezetko and Martin Ehrensvrd and Dating Archaic Biblical Hebrew Poetry: A Critique of the Linguistic Arguments by Robyn C. Vern.

In my view, the essays in Diachrony in Biblical Hebrew (Diachrony in Biblical Hebrew (ed. Miller-Naud and Zevit) successfully answer the arguments of Young et al showing that it is indeed probable that the observable linguistic differences between Classical and Late Biblical Hebrew are due to their date of composition. On the other hand, despite a hyper-critical review by Pat-El and Wilson-Wright (Features of Archaic Biblical Hebrew and the Linguistic Dating Debate), it is quite possible that the conclusion of Vern (quoted below) will be sustained -

No archaic linguistic feature, either singly or in combination across the range of forms, provides evidence relevant for dating the archaic poetry of the Hebrew Bible. The presence of archaisms in the Archaic Biblical Hebrew corpus indicates a poetic style which uses linguistic forms from another period, a common feature of poetry in many cultures.

The ABH poetic corpus is typologically more representative of first millennium sources than second millennium sources. This does not imply that an individual poem cannot be of late second millennium provenance. Up until perhaps 15 years ago it was routinely stated that Biblical Hebrew could be roughly divide into three chronological levels Archaic Poetic (late second to early first millennium BCE), Classical/Early (10th 6th c BCE) and Late (after 6th c BCE) .

It is now clear that much additional work must be done before the usefulness of language analysis in dating biblical passages can be reassessed. This is well described in the last paragraphs of Zevit 2004.

2. Changes in the Pronunciation Tradition of Biblical Hebrew Between EBHP and that Recorded in the Tiberian Masoretic Tradition (early 10th century CE)

Box 10

Justification of Proposals for Early Biblical Hebrew Pronunciation

A written language has no sounds. It does not speak, in a conventional sense, but communicates non-verbally. Language is abstracted into a series of signs that themselves relate information. In writing, language becomes a series of signs.
When a language exists only in written form, the audible sounds that the language contains are not readily apparent. Without the testimony of a native speaker, it may be difficult to match the written sign with specific phones. Discovering the sounds native to a written language, then, is indirect. The internal sound patterns of the language,
external transcriptions in other languages and scripts, and notices about speech behaviour recorded by contemporary witnesses can compensate for the lack of direct phonetic data. From these sources, the phonetic base oa a written language may be recovered [6].


If we assume that the Tiberian Masoretes simply encoded a traditional pronunciation, it is reasonable to insist that any proposals regarding the grammar and pronunciation of EBHP and JEH must be supported by a reconstruction of how the form could have developed into the attested TH given our understanding of the linguistic changes that took place between EBHP/JEH andTH. (Of course, the same requirement separately exists for BHQum, BHPal, and BHGk-Lat)[7].


Tiberian Masoretic Text (MT) has in general satisfactorily preserved the consonantal system of pre-exilic Hebrew. However, it is clear that the vocalization of the MT differs systemically in many ways from the pronunciation of EBHP of over a millennium earlier. These systemic differences, many of which were influenced by Aramaic, can often be identified through comparative grammar. Among the most important changes, mainly phonetic, which can be detected in Hebrew after 600 BCE, are the following. As you will note, some of these changes had already begun to take place before the exile[8].


a) The process whereby the place of stress replaced vowel and consonant length as phonemic went to completion[9]. The Tiberian vocalization system (/TH/+) marked:

  all the phonemes in their reading tradition;

  such allophones (eg. פ = p [f] and gemination) as were required for correct reading of the biblical text according to the Tiberian reading tradition.

The Tiberian system did not explicitly mark vowel length - see Were there Long and Short Vowels in Tiberian Hebrew (TH)?


b) Disappearance of intervocalic /h/.

  This had been well advanced in the pre-exilic period[10]. E.g.

*/lhasˈsuːs/ > /lasˈsuːs/ לסוס <lsws> for the horse[11];

*/yahaˈmiːd/ or */yəhaˈmiːd/ > /yaˈmiːd/ ישׁמיד <ymys> "he will destroy".

  In a few cases it is unknown when the intervocalic /h/ disappeared. The most important case is that of the third person masculine pronominal suffix.

  In the post-exilic period this went further e.g. /lahaˈmiːd/ (/EBHP/); /ləhaˈmid/ (/TH/+);
/laˈmiːd/ לשׁמיד <lhmyd> (MH ) to destroy[12]


c) Elision of syllable-or word-final glottal stop (//[ʔ]) and /y/ usually with a lengthening of the preceding vowel

d) <שׂ> /ś/ [ɬ] > <שׂ, ס> /s/ [s] this commenced before the finalization of the consonantal text of the Hebrew Bible as is shown by a number of cases where original שׂ ś is written ס s. E.g. ספק = שׂפק = to be sufficient etc..

e) The insertion of a short vowel into non word-final diphthongs
בית */ˈbayt/ (/EBHP/) בַּיִת /ˈbayit/ (/TH/+); מות /ˈmawt/ (/EBHP/) /ˈmt/ [ˈmɔːθ] (TH) מָוֶת.[13]

f) 'Segolation'[14]

g) Philippi's law

h) Law of attenuation

i) Spirantization of the bgdkpt Consonants

j) Neutralization of velar and pharyngeal phonemes (//>//, /ġ/>/c/)[15] . This resulted in the elimination of the phonemic distinction between some words. (See Lexicon of Unmarked Consonantal Phonemes in Biblical Hebrew /ġ/[ɣ] AND Lexicon of Unmarked Consonantal Phonemes in Biblical Hebrew /ḫ/ [x])


  עד = as far as - */cad/ (/EBHP/) > /cad/ (/TH/+)

  עד = permanently, forever - */ˈġad/ (/EBHP/+) > /ˈcad/ /TH/+

  חלשׁ <ḥl>. Two distinct roots are found in EBHP which merge when /ḫ/>/ḥ/

     ḥl '"to be weak"

     *ḫl '"to defeat"


k) Pretonic vowel lengthening

l) Reduction of certain vowels to shewa (*/yidˈruū/ (/EBHP/+) /yidrәˈu/ (/TH/+) *[yirəˈʃuː] ([TH]) יִדְרְשׁוּ they sought etc.) or, in the environment of a laryngeal consonant, to another ultra-short vowel (e.g. */yimˈcauː/ Tiberian /yimcăˈu/ (/TH/+) יִמְעֲטוּ)

m) Weakening of the pharyngeal and laryngeal consonants[16] which resulted in:

  The loss of the ability of these consonants to geminate[17] which in turn often caused a lengthening of the preceding vowel[18]. E.g. ברך = he was blessed */burˈrak/ (/EBHP/)
(/TH/+) *[boːˈrɐːx] ([TH]).

  Vowel changes before gutturals (laryngeals)E.gs.

         שמע hearer, hears (ms. qal a.p.) */ōˈmeːc/ (/EBHP/+)
ẹac/[19] *[ ʃoːˈmẹːɐc] (TH). Cf. to the parallel forms in a root identical except that it does not have a guttural - שמע = hearer, hears (ms. qal ap.)
(/EBHP/+) /omẹr/ *[ʃoːmẹːr] (TH).

         שמעת hearer, hears (fs. qal ap.) */ōˈmact/ (/EBHP/+)
*[ʃoːˈmɐː.ɐcθ] (TH). Cf. to the parallel forms in a root identical except that it does not have a guttural - שמר guard, guarding (ms. qal ap.)
(/EBHP/+) /oˈt/ *[ ʃoːˈmɛːθ] (TH).

         At times these changes eliminate important distinctions maintained in pre-exilic Hebrew - e.g. TH qal and hiphil PC 3ms. is יַעֲלֶה while the EBHP would have been - qal */yicˈl/ ; hiphil */yacˈl/.

3. Guidelines I Have Used in Reconstructing the EBHP Vocalization of the First Temple Period Hebrew

(1) Syllables

a. Syllabic Structure [20]

Every syllable in EBHP had one of the following patterns[21] which are similar to some varieties of spoken Arabic[22]:

  CV = consonant short vowel e.g. */l/ "to, for" TH /lə/לְ ;

  CVV = consonant long vowel e.g. /ō/, the first syllable of TH שׁוֹמֵר (*/ōˈmeːr/ (/EBHP/+) );

  CVC = consonant short vowel consonant e.g. /yim/ in יִמְעֲטוּ pre-exilic */yimˈcaū/ > /yimcăˈu/ [yimʕăˈ] (TH);

  CVVC = consonant long vowel OR diphthong consonant e.g. (/EBHP/+)
/ˈsūs/ "horse"; */ˈbayt/ "house"

  CVCC = consonant short vowel consonant consonant e.g.
*/ˈmalk/ (/EBHP/) > /ˈˈlɛk/ [ˈmɛːˈlɛx] (TH). (In TH these mostly developed later into segolates (see http://www.houseofdavid.ca/problem5.pdf) though some final consonantal clusters remain e.g. וַˈיֵּבְךְ ).

From the point of view of syllable length these can be divided into 3 quantities;

  Short Syllables - i.e. CV = consonant short vowel;

  Medium Length Syllables - i.e. CVV = consonant long vowel OR diphthong; or CVC = consonant short vowel consonant;

  Long Syllables - i.e. CVVC = consonant long vowel consonant; or CVCC = consonant short vowel consonant consonant .

Words Significantly Different in Pronunciation in EBHP

Numerals in Pre-Exilic Hebrew


c. Background to Syllabic Stress - (See excursus Evolution of Pronunciation and Stress Patterns )

d. Marking of Syllabic Stress

     I will assume that primary word stress in BH was limited to: (a) verbs and,
(b) nouns (
substantives, adjectives, numbers, and pronouns[23]) in the absolute case. In the transcriptions, the syllable carrying primary word stress are generally in bold with the IPA symbol ˈ preceding the primary stressed syllable;

     All other words (nouns in the construct case and particles[24] - adverbs (including negatives), prepositions, conjunctions etc.)[25] other than mmonosyllabic prepositions and conjunctions (see below) are assumed to carry a secondary stress which I indicate by the IPA symbol ˌ preceding the syllable carrying the secondary stress;

     Mono-syllabic prepositions and conjunctions, almost always connected to the following word in the MT by a maqqeph/makef (מקף) clearly stand midway between inseparable prepositions, which are never stressed, and ordinary nouns in the construct (See Gesenius Hebrew Grammar 16.1) which carry secondary stress. I have assumed that the following, except when they have become independent forms by being combined with prefixes (other than wa- ), carry no stress. In the transcriptions I have replaced the makef by a hyphen.


Table 10

Mono-syllabic Prepositions and Conjunctions Usually Linked to the Following Word in the MT by a maqqeph/makef (מקף)


/EBHP/[26] [EBHP] [27]



il/ [ʔɛl-]



/ʾal/ [ʔɐl-]



im/ [ʔɪm-]



at/ or it/[28] either possibly pronounced [ʔɛt-]

(sign of direct object of verb)


/kul(l)/ [kʊll-]

all of


/min/ [mɪn-]



/cad/ [ʕɐd-]

up to


/cal/ [ʕɐl-]



/pan/ or /pin/ either possibly pronounced [pɛn-]


(2) Phones and Phonemes (see excursus Phonemic Structure of Pre-Exilic, Tiberian and Israeli Hebrew Contrasted; box Phones and Phonemes)

It must be always remembered that:

         phonemic reconstructions, in our case /EBHP/, show the functional structure of the language's sound system while phonetic reconstructions, in our case [EBHP], attempt to represent how it may have sounded;

         the reconstruction of [EBHP] must be largely based on Tiberian pointing, which is mainly phonemic[29], the consonantal (PMT) text, which is phonemic and comparative Semitic linguistics. This necessitates the reconstruction of /EBHP/ which then serves as the base for the reconstruction of [EBHP];

         phonemic reconstructions will always be more certain than phonetic reconstructions. In our case [EBHP] represents one, out of many, possible reconstructions of how /EBHP/ may have sounded. The most important guide in delineating the range of phonetic variation associated with the vowel phonemes are their ranges of values in modern varieties of Arabic (see Aramaic and Arabic as Guides to Reconstructing EBHP ).


a. Consonants