History of the Hebrew Language by David Steinberg

"Common Semitic or Proto-Semitic has three short vowels (§ 10.5): low/open back velar a, high/close front palatal i, and high/close back velar u with strongly rounded lips. It also possesses the three corresponding long vowels ā, ī, ū. Although additional vocalic phonemes have arisen in various Semitic languages, there are no sufficient grounds to suppose that other vowels belong to the original core of the Semitic phonemic system.... However, the realization of the Semitic vowels a, i, u in actual speech can produce other vocalic sounds, mainly in the case of short vowels (cf. §10.11). There is a widespread tendency in Semitic to pronounce high and low vowels, especially when they are unstressed, as mid vowels [e], [ә], [o].... Thus short [i] and [u] tend to become [ә], as in Ethiopic (§21.30), and the same can happen with [a].... Besides, [i] can easily become [e] by lowering the tongue, [u] becomes then [o]. The lack of appropriate vocalic signs, especially for [ә] and [o], does often not allow determining the presence of these vowels in an accurate way, and "e" will then stand for [ә] and "u" for [o] (§21.3). On the other side, a stressed short vowel tends to become long, and its articulation may at the same time be lowered (e.g. i > ī > ē) or raised (e.g. a > ā > ō)[8]. Some of these new vowels may acquire a phonemic status in a determined language."

Quoted from Lipinski 1997 §21.1.

"In numerous Lebanese dialects both vowel quality and quantity are affected by pause, i, u, and a becoming ē, ō and ā (or ä); thus, e.g., in Bišmizzin, contextual byínzil, "he goes down", byúktub "he writes", byíftaḥ "he opens" appear in pause as byínzēl, byúktōb, byíftāḥ."

Quoted from Morag 1989 (p. 102)

In reconstructing the early Semitic and subsequent Hebrew vowel systems it is essential to keep in mind:

1. the distinction at every stage between the probable phonemic structure of the vowel system and the bundle of phones likely to have made up each phoneme. It is probable that the full natural scale of the principal vowel qualities[9] - i, e, a, o, u - would have been heard in the speech of Semitic speakers throughout the centuries though the specific qualities of these vowels is mostly unrecoverable and would, in any case, have varied with time, dialect etc.

2. that Proto-Semitic is thought to have had a similar vowel and stress system to that of Classical Arabic. The written vowel tradition of Classical Arabic recognizes 3 phonemic qualities of vowel each of which has 2 phonemic lengths- i/ī, a/ā, u/ū[10], However, early Greek transcriptions[11] of Arabic names show that Arabic of the period possessed the following vowels i, e, ə, a, o, u.

3. that ancient Semitic languages, and most modern Arabic dialects, phonemically distinguished between short and long vowels. The long vowels were usually quite distinct but the short vowels easily interchanged. To give an example from Egyptian Arabic[12], a language that parallels Ancient Hebrew in numerous ways, there are three short vowels i, a, u and 5 long vowels ī, ē, ā, ō, ū [13]. However, the actualization of /a/ includes [a] and [å]; that of /i/ includes [i] and [e]; that of /u/ includes [u] and [o]. The allophone pronounced depends on such factors as: the nature of the surrounding consonants; whether the syllable is long or short, closed or open; stress; dialect; speed of speaking and even the sex of the speaker[14]. In Palestinian Arabic /u/ is pronounced [o] and /i/ [e] before the (non-geminated) final consonant of words[15]. Thus /ʾuktub/ is pronounced [ʾuktob] and /kātib/ is pronounced [kāteb]. It is interesting to note the similarity of result, regarding the final vowel, to the qal imperfect יקטל (TH יִקְטֹל; EBHP /yiqˈtul/ [yiqˈtul] or [yiqˈtol]), and SC כבד (TH כָּבֵד; EBHP /kaˈbid/ [kaˈbid] or [kaˈbẹd]) and the qal active participle קטל (TH קֹטֵל; EBHP (constr.) [qoːˈtil] or [qoːˈtẹl]).

Unlike the living Arabic dialects, we can never recapture the rich reality of the sound of EBHP. A possible indication of the missing dimensions is given by Rice and Sa'id in their book Eastern Arabic (p. 5) -

In addition to word stress, Arabic also has another system of prominence that works independently of stress. We call this vowel prominence. Like stress, it too is automatic. A long vowel has more sonority (amplitude, loudness) than a short vowel ....

A short vowel immediately followed by a double consonant is more tense than a short vowel elsewhere.... This tenseness is preserved even when the double consonant is not followed immediately by a vowel...

As a result of these three features of word stress, sonority, and tenseness, the acoustic impression of Arabic is quite different from that of English.

Table 17

Long Vowels in EBHP by Origin

	Irreducible Long Vowels
Long Vowel	Primitive Long Vowel	Vowel Lengthened Through Contraction		Vowel Lengthened Through Stress
Long Vowel	Primitive Long Vowel	Homogeneous Diphthong Contraction	Heterogeneous Diphthong Contraction	Vowel Lengthened Through Stress
[iː]	ῑ	íwy >úːy >íyy >î [íː] e.g e.g. /ˈkiwyu/ > /ˈkûy/ > /ˈkiyy/ > /ˈkiy/ > /ˈkî/ (TH כִּי) 'burning' íy > î [íː] e.g e.g. /ˈyíybašu/ > /ˈyîbašu/ > /yîˈbaš/ (TH יִיבַשׁ) 'it will be dry'. Word-final iy > î e.g /ˈkalyu/ > /ˈkaly/ (/EBHP?/) > /ˈkaliy/ (/EBHP?/) > /ˈkalî/ (/EBHP?/) (TH כְּלִי [kәˈliː] (contextual) or כֶּ*לִי [ˈkɛːli] (pausal) 'tool' /ˈyihyay/ > /ˈyihy/ (/EBHP?/) > /ˈyihiy/ (/EBHP?/) > /ˈyihî/ (/EBHP?/) > (TH /yˈhi/) "may he be" úy >* î e.g. /wayˈyúyśam/ > /wayˈyîśim/ (TH וַיִּשֶֺם ) 'he put'.
[eː]	TH מֵת - see next column.	Word-final áyu > ê [éː] e.g. /šamōˈnayu/ (PH) → (/EBHP/) /šạmōˈnê/ > (TH) /šmōˈnê/ Word-final íyu > ê [éː] e.g. /bāˈniyu/ → /bōˈnê/ ('building' qal a.p. ms.) מת (TH מֵת) 'dead' (adj.) - the origin of the long eː is unclear i.e. it might have been /ˈmêt/ or /ˈmēt/. In either case we should see מֵת the 3ms. SC as having a stress lengthened, and hence reducible, vowel i.e. /ˈmeːt/. In transcriptions of EBHP I will use /ˈmêt/ qiʾl > qệl > qʾệl	Unstressed diphthong contracts ay > ê [eː] e.g. */bayt/ > /bêt/ (TH בֵּית) 'house of-'	/kaˈbidu/ > /kaˈbeːd/ (TH כָּבֵד) 'heavy' (adj.)
[aː]	TH קָם - see next column.	Word-final áya > [áː] e.g. /baˈnaya/ (PH) > /baˈnay/ > /baˈnâ/ (/EBHP/) קם (TH קָם) 'standing' (qal ms. ap.) as an example of the large class of II-w and II-y roots. The origin of the long aː is unclear i.e. it might have been /ˈqâm/ or /ˈqām/. In either case we should see קָם the 3ms. SC as having a stress lengthened, and hence reducible, vowel i.e. /ˈqaːm/. In transcriptions of EBHP I will use /ˈqâm/ for the participle and */ˈqaːm/ for the 3ms. SC.		/ˈyadu/ > /ˈyaːd/ (TH יָד) 'hand'
[oː]	*ō (ō < ā)* e.g. /ˈqāṭilu/ (PNWS) → /qōˈṭeːl/ (/EBHP/)	בוש (TH בּוֹש) 'ashamed' (adj.) - the origin of the long oː is unclear. /ˈbôš/ áʾ not immediately followed by a vowel shifts to â [áː] e.g /ˈraʾ*šu/ > /ˈrâšu/ → /ˈrôš/ (/EBHP/) 'head'.	Unstressed diphthong contracts aw > ô [oː] e.g. */mawt/ > /môt/ (TH מוֹת) 'death of-'	/gaˈdulu/ > /gaˈdoːl/ (TH גָּדוֹל) 'big' (adj.)
[uː]	ū	úw > û e.g. /ˈhúwšabtima/ → /hûšabˈtim/ e.g (TH הושַבְתֶּם) 'you were made to dwell' íw > û e.g. /ˈyíwkalu/ > /ˈyûkalu/ → /yûˈkal/ (TH יָכוֹל ) 'he will be able'.

4. that while there two phonological vowel lengths there often (always?) are at least 4 phonetic vowel lengths i.e.

o short vowels are longer when stressed particularly in closed syllables. In addition word final short vowels, as often in Arabic^{^[16]}, were probably shortened long vowels in quality rather than lengthened short vowels. Thus for /i/, the short vowel within the word was likely pronounced as the laxer vowel [ɪ] while the short vowel at end of word as the tenser vowel [i]. Cf. to the parallel long vowel pronounced [iː].

o long vowels e.g. /iː/ are longer when stressed [iːˑ].

In the history of Hebrew prior to the middle ages it seems to me that the appearance of new long phonemic vowels may have stimulated a reanalysis of the short vowels to parallel the long vowels[17].

Table 18

Shifts in Proto-Semitic Vowels as Hebrew Developed

Nb. form in [square brackets] is a reconstructed pronunciation of the preceding /phonemic/ form where likely alternatives exist

PS /PS/ (c. 3000 BCE)*	EBHP /EBHP/ (c. 850-550 BCE)*	PH (c. 1200 BCE)*	EBHP /EBHP/ [EBHP][18] (c. 850-550 BCE)	PTH **/PTH/⁺ [PTH] (c. 400 CE)**	TH /TH/⁺ *[TH] (c. 850 CE)	IH [IH] (present)
/a/	a	a - /šaˈlaːmu/ “peace”	a - /šaˈlōm/	/aː/ - /šaːˈlōm/	/å/ - ׄשָלוֹם /šåˈlom/ [šɔːˈloːm]	a - [šaˈlom]
	a	a - /qaˈtalat/ “she killed”	/aː/ /qaˈtala(ː)h/ (/EBHP?/) > /qaˈtalaː/ (/EBHP?/)	/aː/ - /qaˈtalaː/ > /qaːˈtalaː/ > /qaːtәˈlaː/	/å/ - קָטְלָה /qåṭәˈlå/ [qɔːṭәˈlɔː]	a - [katˈla]
	a-a	a - /ˈdabaru/	a - /daˈbaːr/	/aː/ː - /daːˈbaːr/	/å/ - דָּבָר /dåˈbår/ [dɔːˈvɔːr]	a - [daˈvar]
	a	a - /ˈqatalat/ “she killed”	a - /qaˈtala(ː)h/ (/EBHP?/) > /qaˈtalaː/ (/EBHP?/)	/ә/ - /qaˈtalâ/ > /qaːˈtalâ/ > /qaːtәˈlaː/	/ә/ - קָטְלָה /qåṭәˈlå/ [qɔːṭәˈlɔː]	∅ - [katˈla]
	a	a - /ˈkattaba/ (piel perf. 3rd m)	i - /kitˈtib/ [kitˈtiˑb] or [kitˈteˑb]	i - /kitˈtẹb/	i - כִּתֵּב /kitˈtẹb/ [kitˈtẹːˑv]	i - [kiˈtɛv]
	i	a - /ˈkattaba/ (piel perf. 3rd m)	i - /kitˈtib/ [kitˈtiˑb]/[kitˈteˑb]	e - כִּתֵּב /kitˈteb/	ẹ - כִּתֵּב /kitˈtẹb/ [kitˈtẹːˑv]	ɛ - [kiˈtɛv]
	a	(particle attached to the direct object)	a - /ˌat/ [ʾat] or [ʾet]	e - /ʾat/ > /ˌʾet/	ẹ - אֵת /ˌʾẹt/	[ɛt] or [t]
	a	(particle attached to the direct object)	a- /ʾat/ [ʾat]/[ʾet]	e - /ʾat/ > /ʾet/	ɛ - אֶת־ /ʾɛt/	[ɛt] or [t]
/aː/	/oː/	/aː/ - /šaˈlaːmu/ “peace”	/oː/ - /šaˈloːm/	/oː/ - /šaːˈlōm/	/oː/ ׄשָלוֹם /šåˈlom/ [šɔːˈloːm]	a - [šaˈlom]
/aː/	ō	/aː/ - /ˈṭābu/ “good”	/ō/ - /ˈṭōb/	/ō/ - /ˈṭōb/	/o/ - /ˈṭob/ טוֹב [ˈṭoːˑv]	o - [ˈtov]
/i/	i	/i/ - /ṣiˈrāru/ “bag”	/i/ - /ṣiˈrōr/	ә - /ṣәˈrōr/	ә - צְרוֹר /ṣˈror/ [ṣәˈroːr]	∅ - [tsˈror]
	i	/i/ - /ˈsiprahu/ “his book”	/i/ - /sipˈrahu/ > /sipˈrô/	/i/ - /sipˈrô/	/i/ - סִפרוֹ /sipˈro/ [sifˈroː]	i - [sifˈro]
	i	/i/ - /i/ /qitˈtil/ (adjectives of infirmities)	/i/ - /i/ - /ʾilˈleːm/ "deaf"	i -e - /ʾilˈleːm/	i - ẹ - אִלֵּם /ʾilˈlẹm/ [ʾilˈlẹːm]	i - ɛ - [iˈlɛm]
		/i/ - /i/ /qitˈtil/ (adjectives of infirmities)	/^ciwˈweːr/ "blind"	/^ciwˈweːr/	עִוֵּר /^ciwˈwẹr/ [^ciwˈwẹːˑr]	[iˈvɛr]
	i	i - /ˈʾāsiru/ (qal ac. participle) “one who ties”	ˈi - /’ōˈseːr/	ˈe - /’ōˈseːr/	ˈẹ - אׄסֵר /’oˈsẹr/ [’oːˈsẹːr]	ɛ -- [oˈsɛr]
	i	i - /ˈʾilu/ “god”	i - /ˈʾeːl/	e - /ˈʾeːl/	ẹ - אֵל /ˈʾẹl/ [ˈʾẹːl]	ɛ - [ˈɛl]
	i	i - /ʾilu/ “god of”	i - /ʾil/	e - /ʾel/	ɛ - אֶל /ʾɛl/ [ʾɛl]	ɛ - [ɛl]
/ī/		/iː/ - /’aˈsīru/ “prisoner”	/iː/ - /’aˈsiːr/	iː - /’aːˈsiːr/	ˈi - אָסִיר /’åˈsir/ [’ɔːˈsiːr]	ˈi - [aˈsir]
/ī/		/iː/ - /ˈmīnu/ “kind, variety”	ˈī - /ˈmīn/	iː - /ˈmīn/	ˈi - מיִן /ˈmin/ [ˈmiːn]	ˈi - [ˈmin]
/u/		/u/ - /ˈdubbara/ “it was said”	u - /dubˈbar/	u - /dubˈbar/	u - דֻּבַּר /dubˈbar/ [dubˈbar]	u - [duˈbar]
		/u/ - /ˈruḥābu/ > /ruˈḥōbu/ "plaza inside city gate"	u - /ruˈḥōb/	ә - /rәˈḥōb/	ә - רְחׂב /rˈḥob/ [rәˈḥoːv]	ә - [rәˈxov]
		/u/ - /ˈgudlahu/ “his greatness”	u - /gudˈlahu/ > /gudˈlô/	o - /gudˈlô/	ɔ - גָּדלוֹ /gǫdˈlo/ [gɔðˈloː]	o - [godˈlo]
		/u/ - /ˈyaqum/ “let him stand”	/u/ - /ˈyaqum/	u - /yaːˈqum/	/o/- יָקׄם /yåˈqom/ *[yɔːˈqoːm]	o - [yaˈkom]
		/u/ - /wayˈyaqum/ “he stood”	/u/ - /wayˈyaqum/	u - /wayˈyaːqum/	/ɔ/ - /wayˈyåqǫm/ *[wayˈyɔːqɔm]	[vaˈyakom] or [vayaˈkam]
/ū/		/ū/ - /yaˈqūmu/ “he will stand”	/uː/ - /yaˈqūm/	uː - /yaːˈqūm/	/u/ - יָקוּם /yåˈqum/ *[yɔːˈquːm]	u - [yaˈkum]
/ū/		/uː/ - /ˈṭūbu/ “goodness”	/uː/ - /ˈṭūb/	uː - /ˈṭūb/	/u/ - טוּב /ˈṭub/ [ˈṭuːv]	u - [ˈtuv]
/aw/[19]		áw - /ˈmawtu/ “death”	áw - /ˈmawt/	áːwe - /ˈmaːwet/	ˈåwe - מָוֶת /ˈmåwɛt/ [ˈmɔːwɛθ]	áve - [ˈmavɛt]
/aw/[19]		aw - /mawt/ “death of”	aw - /ˌmawt/ [ˌmawt]/[ˌmɔwt]	oː - /ˌmôt/	o - מוֹת /ˌmot/ [ˌmoːθ]	o - [mot]
/ay/		áy - /ˈbaytu/ "house"	áy - /ˈbayt/