Edition 2.0

12 December 2011


E-book - History of the Ancient and Modern Hebrew Language

E-book - Biblical Hebrew Poetry and Word Play - Reconstructing the Original Oral, Aural and Visual Experience

By David Steinberg


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


To print use PDF file here


Table 8

Changes in Pronunciation Between EBHP, TH and BHIH

most of which Alter the Syllabic Structure of Nouns




(c. 1200 BCE)


*/EBHP/+ *[EBHP][1]

(c. 850-550 BCE)


/TH/+ *[TH]

(c. 850 CE)

[BHIH] = [IH]


Segolate Masculine singular

ילד "boy"

/ˈwaldu/ > /ˈyaldu/


[ˈyɐld]? [ˈyɐlәd]?

(phonemically 1 syllable)



(2 syllables)


(2 syllables)

גבר "man"




(phonemically 1 syllable)



(2 syllables)


(2 syllables)

ספר "book"




(phonemically 1 syllable)



(2 syllables)


 (2 syllables)

קדש "holyness"

/ˈqudu /



(phonemically 1 syllable)



(2 syllables)



מות "death"







מות "death of-"







Segolate Masculine Plural

מלכים "kings"








"kings of-"



[mɐl(ɐ)ˌkɐy]? [mɐlɐˌkɛy]?




Segolate masculine singular and addition of short helping vowel to ease pronunciation of guttural ח

תחת "beneath, instead of-"

(originally substantive functioning as preposition)



[ˌtɐħt]? [ˌtɐħәt]?

(phonemically 1 syllable)



(2 syllables)


(2 syllables)

פחד "fear, trembling"





(phonemically 1 syllable)



(2 syllables)


(2 syllables)

noun ending

עפרת "lead"







(phonemically 2 syllables)



(3 syllables)


(3 syllables)

פלשת 'Philistia"

/paˈlitu/ >





(phonemically 2 syllables)



(3 syllables)


(3 syllables)

Other Segolate

פרי "fruit"


/ˈpiryu/ > /ˈpiry/
[ˈpɪɾy]? [ˈpɪɾĭy]?


(pausal /ˈpɛri/

[ˈpɛːɾiː] )

(pausal [ˈpɛʁ̞i])

Synope of Certain Reduced Vowels[4]

אדמתי "my earth"







Post First Temple Period addition of short helping vowels to ease pronunciation of gutturals this has the effect of adding a syllable

גבה "high" (adjective)

/gaˈbuhu/ >



(h is consonantal)

(2 syllables)



(h is consonantal)

(2 syllables)


(3 syllables)

יעשה "he will do"


/yacˈś/ >


(2 syllables)



(3 syllables)


(3 syllables)

שמח "happy"




(2 syllables)



(2 syllables)


(3 syllables)

רקיע "firmament"




(2 syllables)



(2 syllables)


(3 syllables)

נח "Noah"




(phonemically 1 syllable)



(1 syllable)


(2 syllables)

יעקב "Jacob"




(2 syllables)



(3 syllables)


(2 syllables)

Heterogeneous Diphthongs

מים "water"

/ˈmaymu/ >
/ˈmaym/ like the English word 'mime'



(1 syllable)



(2 syllables)


(2 syllables)

שמים "heavens"




(2 syllables)



(3 syllables)

(3 syllables)

מצרים "Egypt"







אפרים "Ephriam"








בית "house"







ירושלם "Jerusalem"











[bayt] > [bayit]

Quoted from Development of the Canaanite Dialects: An Investigation in Linguistic History by Zellig S. Harris, American Oriental Series, Vol 16, 1939 pp. 29-32

[ay] > [ȇ]; [aw] > [ȏ]

In early Semitic, diphthongs were phonologically vowel + syllable-closing [y] or [w] ; as such they were always either final or followed by the consonant which began the next syllable : [baytu]. Since every syllable in early Semitic began with a consonant, inter-vocalic [y] and [w] must be considered phonologically as hetero-syllabic, not making a diphthong, but rather beginning the next syllable: [baytiya] of my house.' In Canaanite, diphthongs were monophthongized in all positions, accented and unaccented, medial and final, except when another [y] or [w] followed; thus [ḥayym] life,' [ḥayyȇ] `life (cst.),' [taw] < [tawwu] mark,' all remained in Phoenician and Hebrew. The diphthongs [iy], [uw] in medial position had been monophthongized. in Canaanite times.

New diphthongs arose later in final position, all of which, with the exception of the last group, were later monophthongized when [h] elided in the 3rd person suffixes in Hebrew [-ahū] > [-aw], somewhat later > [-ȏ] .

In Jerusalem Hebrew, unaccented diphthongs were monophthongized as throughout Canaanite, but accented diphthongs remained. In the nifal verb, the monophthongization could not take place until the verb stress shift : [nawdaca] > [nawdac] > [nawdac] > [nȏdac] it became known.' This was also true of those hifil verb forms which did not have [] in the second syllable, e. g., the infinitive absolute: the Assyrian transcription a-u-si-' (III Rawlinson 10. 2. 28) [hawic], later > [hȏȇac], shows the form before the verb stress shift, or in any event before the reduction of the diphthong. This will explain the biblical Hebrew forms, except for the absolute nouns of type [ḥȇq] bosom,' [yȏm] day,' which may be due to analogy of the construct forms, or to borrowing from Hebrew dialects where the monophthongization had been complete (cf. the possibly dialectal [lȇl] in Jes. 21.11, variant to [layl] night.'

Later, after final short vowels were dropped, and the medial diphthongs came to be in doubly closed syllables, they were pronounced as two syllables; [bayt] > [bayit]; [cȇnaym] > [cȇnayim]; [mawt] > [mawɛt]. This was part of the late general tendency to break final consonant clusters by anaptyctic ("segolate") vowels. Final diphthongs remained: [mātay] when.'

It has been suggested that this divergence of Jerusalem Hebrew is a later development, that Jerusalem had originally gone with the rest of Canaanite, but that later foreign influences caused a restoration of the diphthongs in some cases. Such new formations, extending from loan-words which might have come to Jerusalem from a dialect where diphthongs had been preserved, would indeed be possible. However, the fact that the diphthong does not occur in some special group of words or in some morphologic class, but can be explained as having been preserved in one phonetic environment (stress), argues for a regular phonetic development. The probability is therefore that when this change first spread in Canaanite there were some areas, specifically Jerusalem, in which stress was a deciding factor for its occurrence.


[1] Note, in reconstructed [EBHP] transliterations and sound files -

1.there is no spirantization of the bgdkpt consonants;

2. vowel qualities are outlined here;

3. I use the most probable form. Where no one form stands out as most probable, I select the one closest to the MT vocalization.

4. when multiple forms are possible, the form used is underlined.

[2] See Harris 1939 p. 67.

[3] This reconstruction is directly verified by Khan 1990 p. 97.

[4] See Harris 1939 p. 67.