Like the w:Formosan languages which had fanned out from Taiwan ca. 3000 BCE, an intense flattening of the phonemic landscape slowly decimated the Semitic language(s) in Oceania. By the time of the writings we have from w:Mindoro, many simplifications had taken place. Almost all the changes parallel the changes in Proto-Austronesian (PAn) > Proto-Malayo-Polynesian (PMP) and subsequent developments.
Any attempt to reconstruct the sound of AH is hampered by several causes. First, we have no unambiguous record of the sound of Hebrew in the ANE ca. 1000 BCE. The MT is from many centuries later, and the few inscription we possess are consonantal scripts. Second, the development of writing in Southeast Asia did not catch hold for nearly a millennium after the Semitic people had departed into Polynesian.
The following is a reconstruction of Early Biblical Hebrew phonology, perhaps as late as 500 BCE. This understanding is based on Greek transliterations, cross-linguistic evidence, and the MT.
|Early Biblical Hebrew Vowels|
|High||i and iː||uː and ʊ|
|Low||a and aː|
|Early Biblical Hebrew Consonants|
|Best-Guess PH Consonants|
Paleo-Hebrew - the direct ancestor of AH - did not yet have lenition, as found in Aramaic. Students of Hebrew use the acronym w:begadkefat to help remember that the presence of a strong dagesh means not to lenite those stops into fricatives. The sin-šin distinction may have begun, but was not marked in the orthography. Two letters (ayin and het) were homographs., and so the distinction between uvular and pharyngeal fricatives is sometimes hard to follow.
Contrastingly, the vowels of Paleo-Hebrew seem relatively close to Proto-Semitic (which had only /a i u/ and did not contrast length). AH seems to have evolved along similar lines to it's siblings back in the ANE and to w:Chamorro language and w:Palauan language. The PAn schwa probably had a AH reflex of /ε/. A system of five vowels had emerged, with all possibilities contrastive for length.
The amalgamated Semitic people who found themselves sold into Southeast Asia had a phonology that largely overlapped with the surrounding PAn language(s). The sounds which were uniquely Afro-Asiatic seemed to have dropped off very quickly. The most startling change was PH *ɬ > AH *ŋ. The addition of the velar nasal may seem strange, especially in the syllable onset, but is entirely predictable given the new surroundings in Southeast Asia and Oceania. The alveolar fricative ejective (Tiberian צ/ṣaḏé) may well have become the alveolar affricative before leaving the Levant.
- PH *k' > AH q
- PH *χ > AH q
- PH *ʁ > AH r
- PH *t’ > AH t͡s
- PH *ʃ > *ɬ > AH ŋ
- PH *ɬ/*ɬʼ > AH ŋ
- PH *ħ > AH h
- PH *ʕ > AH ʔ
- PH *θ > AH t
- PH *z > AH ɟ͡ʝ
- PH *s’ > AH t͡s
- PH *ʔ > AH zero, except word-initially, where it persisted
After untold hundreds of year surrounded by and serving in the Balangays round about them, the Hebrew language had grown to sound like it neighbors in almost every respect. Still like the ANE, n assimilated causing gemination before everything except ʔ , h, r and (innovatively) q.
ɟ͡ʝ is written z and t͡s is written c.
|High||i vs. iː||u vs. uː|
|High-mid||e vs. εː||o or oː|
|Near-low||a vs. aː|
The semitic phonotactics remained largely in place during this period, with CV, CVC, and CV1V1C allowed. However, the contraction of diphthongs began to stop, and so the presence of glides produced more complicated syllables. We should, therefore, expand our notation to include CVG and CVGC, or more conventionally CV1V2 and CV1V2C.
Semitic phonotactics required every syllable to begin with a consonant, but the disappearance of certain consonants medially and finally allowed vowel hiatus for the first time. The seeming anti-diphthong bias of the Levant was replaced by the pro-diphthong attitude of Polynesia.
Short vowels in an open, unstressed, pretonic syllables reduced (i.e. elided) when possible, just as in Amorite and Ugaritic. Short, stressed vowels often lengthened, as in Phoenician.
- i + y = î ; e + y = ey/'ē ; a + y = ay ; o + y = oy ; u + y = uy/'ū
- i + w = û ; e + w = û ; a + w = ô ; o + w = ow ; u + w = û
- a + o = ô ; e + o = ô ; i + o = ô ; u + o = ô
w + a = ā ; e + a = ê/â ; a + e = ê/â ; o + a = â
While nothing like Akkadian's full-blown vowel harmony occurred, there were definitely "e-colored" and "a-colored" environments, that is times when E won and visa versa.
The 'begadkafat letters lost control of their voicing in the coda, but this was not noted in the orthography.
|"Gutterals"||"Sonorants"||← or →||"Begadkap̅at"|
|able to be...||h||'||q||r||w||y||l||m||n||ŋ||s||c||z||b||g||d||k||p||t|
|geminated||No||No||No||No||as uw||as iy||Yes||Yes||Yes||Yes||Yes||as t||Yes||Yes||Yes||Yes||Yes||Yes||Yes|
|in coda||Yes||No||No||Yes||as u||as i||Yes||Yes||Yes||Yes||Yes||as s||as d/r/s||w.u.v.||w.u.v.||w.u.v.||w.u.v.||w.u.v.||w.u.v.|
- JBL 124, No. 2, Richard C. Steiner, p.229-267
- likely akin to PAn *ɬ > PMP *ŋ, l, n.
- See geographic distribution, The World Atlas of Language Structures Online, Chapter 9: The Velar Nasal
- The emphatic consonants are hotly debated, so /k'/ may well have been /q/ in the ANE.
- Likely already done in the ANE.
- That's not entirely true, since /u/ aka šureq had always existed