Hebrew Phonology

Consonants

¹ Consonants / ħ / and / ʕ / are pronounced only by elderly people from Arab countries. Most carriers use / x / and / ʔ / instead.

² Consonants / ʔ / and / h / are regularly pronounced only in formal speech, but in the spoken language they most often fall out.

³ Usually transcribed as / r / . Pronounced as / ɣ / , less often as ʁ , can sometimes be pronounced as / ʀ / or / r / , depending on the medium.

⁴ Sound / n / is pronounced as / ŋ / before velor ^[1] .

Noisy consonants can be assimilated by voicing. Deaf / pt ts tʃ kfs ʃ x / go into voiced / bd dz dʒ ɡ, vz ʒ ɣ / when they are directly in front of the voiced noisy, and vice versa. Examples:

• לסגור / lisˈɡor / > / lizˈɡor / ('close'), / s / > / z /
• זכות / zxut / > / sxut / ('right'), / z / > / s /
• חשבון / xeʃˈbon / > / xeʒˈbon / ('score'), / ʃ / > / ʒ /
• מדפסת / madˈpeset / > / matˈpeset / ('printer'), / d / > / t /
• אבטחה / avtaˈxa / > / aftaˈxa / ('security'), / v / > / f /

In modern Hebrew, alternations b ~ v , k ~ x and p ~ f are permissible (in each pair the first variant is used after the consonants and in the place of the historical double consonants). In ancient Hebrew also t ~ θ , d ~ ð and ɡ ~ ɣ .

Vowels

Hebrew vowels are usually divided according to Masoretic signs for long, short and super-short vowels (except for always long / o / and super-short seams). Superfine are not / i / and / u /. In modern speech, historical vowels merged into a simple pattern of five phonemes. There are diphthongs / aj / and / ej /.

The historical seam in modern Hebrew is either not pronounced, or pronounced as / e /. Any vowel of modern Hebrew can be reduced to / ə /, when it is far from the stress ^[1] .

Emphasis

In primordial words stress is on the last syllable (most often) or on the penultimate syllable (for example, in the group of names with / e / in the last syllable, called segolatami). In borrowings and primordial neologisms, the stress may not obey these rules (for example, אֵיכְשֶׁהוּ / ˈexʃehu / “somehow”, פּוֹלִיטִיקָה / poˈlitika / “politics”) ^[2] . In colloquial speech, especially in proper names, the stress can shift from the last syllable to the penultimate ^[3] .

Conversation articulation

A vowel that is removed by two syllables from stress can be reduced or disappear: ^[1]

* zót emérət > stemérət ('this means')

* éx korím láx > əkorímlax ('what's your name?')

The sound / l / may fall out after an unstressed vowel, sometimes with the next vowel: ^[1]

* ába ʃelaxém > ábaʃxem ('your father')

* ú itén lexá > uiténxa ('he gives you')

The syllable of type / rV / disappears before / x / , except for the case when this combination appears at the end of the speech flow: ^[1]

* bé-dérex klál > bədéxklal ('usually')

but: ú badé re x ('he is on the road') at the end of the stream of speech, before a pause.

Clusters of t and d turn into one consonant (if they do not go at the end of the speech flow):

* aní lamád-ti páːm > əniləmátipaːm ('I once taught')

but: ʃe-lamá dət i ('what I taught')

The law of the proclamation of syllables

In Hebrew, only five kinds of syllables are permissible, and two kinds are never unstressed and one is never stressed. Most of the changes in the structure of words occur for the sake of adhering to the law of vocalization of syllables. Under the influence of guttural sounds and ר , changes occur in the words, which can sometimes be regarded as a violation of the law of syllable voicing.

All syllables begin with a consonant (including alef, guttural bow), except for one case: the conjunction ו “ and” before the labial sounds ב , מ, פ is pronounced as ū.

The emphasis falls on the last syllable, sometimes (in segolaty and quasi-seals) - on the penultimate.

The table according to the law of syllable vocabulary is presented below (C - consonant - consonant; V - vowel - vowel) ^[4] .

The case of two syllables of type C

If two letters from the seam follow in succession, the first one is at rest and should be part of another syllable (close it), and the second one is movable and forms its own syllable of type C. If a situation occurs in which both sutures must be mobile (for example , two stitches under the first two letters of the word), then the first stitch turns into i (sometimes, if this stitch is the result of a reduction, it turns into that vowel that was before the reduction). If in a syllable like C instead of a suture there is a hataf (super-short vowel), then in such a situation it turns into a simple short vowel (two points disappear graphically). All this is a consequence of the law of syllable vocation, according to which two syllables of type C cannot follow in succession ^[5] .

In a situation where, under the first letter of the word, a moving seam, and under the second, as a result of the reduction, the vocalization turned into a seam, the first seam turns into another vowel (as described above), and the second (called “fluttering”) does not pronounce and closes the syllable (like CVC). The peculiarity of the flashing seam is the absence of light Dagesh after it in the letters of the Beghed Kefet group.

Reduction and reduction

When joining endings or in other cases when the stress changes its position, the stressed syllable that has become unstressed changes.

In closed percussion syllables, long ā, ō, ē become brief ( שומֵר šōmḗr / shomer "guardian, guard"> שומֶרךָ šōmerħā / shomerha "your guardian"). If ō or ē are in front of a double consonant, then they turn into u or i, respectively ( ōב dōv / dov "bear"> דֻבּים dubbīm / dub "bears", חֵץ xēsˤ / hets "arrow" > חִצּים xissˤīm / Hitsim "arrows"). This phenomenon is called vowel reduction. In modern Hebrew, after the disappearance of differences in longitude, only the transition ō / ē to u / i is reflected in the pronunciation.

In the open former stressed syllables, long ā, ō, ē and a short “a” turn into a movable suture ( כתַב kāθav / katav “he wrote”> כתְבו kaθәvū / katvu “they wrote”). Long ā can be reduced in the pre-predardary syllable, which in the original form of the word was pre- shock ( כָתב kāθáv / katav "he wrote"> כְתבתם kәθavtém / ktavtem "you (the husband) wrote"). This phenomenon is called vowel reduction.

In some types of names, as a result of the reduction, a combination of a letter with a small vogue and a letter with Dagesh is formed, under which there is a newly formed movable suture. In such cases, the seam becomes quiescent (fluttering), and the dagesh disappears.

Dagesh

There are two types of daghese . Both are graphically indicated by a dot inside the letter. Laryngeal letters do not take Dagesh.

Heavy (strong) dagesh - historically doubling of the letter. He is not at the beginning of a word. When it is considered a letter for two: the first part (it is believed that it does not have a vow) closes the previous syllable, the second part opens the next.

Light (weak) Dagesh - put in letters ב , ג, ד, כ, פ, ת ( group "beɣeð-kefeθ / runed-kefet"), when they open the syllable after the closed (go after the vocalization of the seam) and the beginning of the word. In antiquity, it designated explosive pronunciation, unlike slit in pairs b / v, g / ɣ, d / ð, k / ħ, p / f, t / θ. In modern Hebrew, the difference is preserved b / c, c / i, p / f, while the rest of the letters have only explosive versions of d, d, t. The light dagesh is not considered a doubling and does not affect the accounting of syllables. With heavy Dagesche, an explosive version of pronunciation is also used (doubled in ancient times).

Filippi law

is a phonetic rule established by F. V.M. Philippi with reference to the Masoretic vocation of the text of the Tanakh. This law states that in the stressed syllable the historical vowel i becomes a. Some researchers consider this law not Hebrew, but Aramaic, associating its appearance in the speech of the Masorets with the Aramaic influence ^[6] .

The origin of this phenomenon is controversial. Apparently, it appeared rather late and not in all dialects. Thus, in the work “Hexaplas”, Origen (the beginning of the 3rd century AD) gives the transcriptions ελλελθ (“you praised”, masor. Hilaltā), εκσερθ (“you cut”, masor. Hiqsˤartā) ^[7] . In the Septuagint (200 BC) the prerequisites for the action of the Philippi law (for example, Sept. [Gεt], Masor. [Gat]) already appear.

The reverse law says that the historical "a" in a closed unstressed syllable goes into "i" (* maɣdal> miɣdal "tower", * šamšōn> šimšōn "Samson"). The origin of this rule dates from 400 to 850 AD. e. Even in the Septuagint and the New Testament, forms with “a” were fixed (in proper names, like “Samson”, “Magdalen”) ^[8] .

Gzarot

The word "ghizra" (pl.-gzarot) refers to phenomena in which the structure of the word-formation model changes under the influence of the letters of the root. In many cases, gzarot are irregular trends that do not always appear.

To indicate irregularity in the root, letters from the root פעל with Gershaim (double apostrophe) or Roman numerals with a hyphen are used, at which stand: the letters ר ( from the word גרוני “ guttural”); the letters א , י, ו, נ \ ן to indicate their presence. For example: פ"גר and I-gtr. - the first letter of the root laryngeal; "י and II- י - the second letter of the root י ( "empty root"). The combination ע"ע means that the second and third letters of the root are the same, and the combination ל"ה is equal to ל"י (in some positions י changes to the vowel ā, denoted by ה) .

Laryngeal letters

The term “guttural letters” refers to letters denoting guttural sounds א , ח, ה, ע . They also include the letter ר, since with it, as with laryngeal, there is no Dagesh, but it does not possess the other properties. Different guttural letters appear to varying degrees in gzarot. Their main properties are:

1. Vowels of the upper rise when guttural have a tendency to turn into vowels of the middle and lower rise;
2. The seam (any) goes into super-short vowels (hatafs). At the same time, a hataf does not happen at the end of a word, and a laryngeal can terminate a word without a vowel. In the letters of the group, the begat-kefet, after the hatafs, is not an easy Dagesh;
3. Throat can not double (take dagesh).

Lower vowels

1. Laryngeal (not resolved) with hirik (i) in front of the letter with the seam at rest. There is a decrease - Hirik is replaced by a segol (e) or patas (a).
2. Throat (unresolved and rarely alef) with segolem (e) after the letter with the shock ē, e or ō. Segoli (e) are transformed into patachas (a).
3. Laryngeal (not decisive) second radically. In the forms of the imperfect and imperative of the breed, pāʕal decreases to a.
4. Laryngeal (not decisive) second at the root. When joining the endings ךָ, כֶם, כֶן , two movable sutures may appear in the names (the first is from the reduction, the second from the endings), then the first suture under the laryngeal is replaced by a.
5. Larynx (not deciding and Aleph) at the end of a word. Small voices, ō in the imperfect and imperative 1 of the breed and ē in verbs are replaced by patas.
6. Larynx (not deciding and Aleph) at the end of a word. After large full voices, incomplete Holam (except in the case above), ē in the names and ē in the verbs (along with the case above), the unstressed “lurking pats” written under the laryngeal is inserted.

Laryngeal and suture

1. Laryngeal (not decisive) with quiescent suture after letter with unstressed a, e, o, i, u. The seam turns into a hataf corresponding to the previous vowel, and with u and i the vowel is replaced by o or a / e and the laryngeal is announced by the hataf. Ghizra does not work if the syllable is “pregortalized” syllable is stressed or was stressed. Guizra always works with alef, it may not work with Aine and Hey, often with a het it does not work or it works half (vowels i and u are lowered, but the seam remains). A hataf before a consonant with a movable suture is not considered a hataf, but is considered a full-fledged vowel (2 points disappear graphically). Khataf-patah before guttural also turns into patas.
2. The movable suture under the laryngeal (not resolved) is replaced by the hataf. Most often it is hataf-patah. When the “qṓtˤel” model changes under the first root, a movable suture may occur, and if it is laryngeal, then the suture goes into the hataf-kamats. In the imperative and inclined infinitive without the preposition of the pāʕal breed, with the first root laryngeal suture goes into hataf-segol under alef and hataf-patah under other laryngeal. In the rare case when alef is at the end of a word after a seam, gizra does not work.

Throat and Dagesh

If the laryngeal (including resh) must have a strong dagesh, it does not accept it, compensating for this by lengthening the vowel preceding it (with i> ē, u> ō, a> ā, and the vowel is unstressed). The vowels appearing as a result of this gizra are not reduced or reduced, even when they are at the right distance from the stress. Compensation always works in the first and third letters of the root. With the second letter of the root, if it is a resh or an alef, it usually works, with a hat it does not work (there is no Dagesh at the same time), with hey and 'ain - it works with the promulgation of Kubbuts, with others it does not work.

Vowel dissimilation

1. In the combination “a letter with patakh (a) + a letter with dagesh with kamats (ā)”, if the second letter is laryngeal, the dagesh is lost, and the pats change to segol (e). When you change the word, if Kamac disappears, the stalemate remains in its place.
2. If, in order to compensate for the dagesh in the perfect breed, pi'el i changes to ē, the second ē can go to a. This phenomenon is observed only in some verbs, all of them have the option without it.

Weak letters

Weak letters are called א , י, ו . How weak can it be נ. Their feature is the possibility of loss, assimilation, alternation. א sometimes shows properties of guttural, sometimes - yud. Weak letters can alternate in different models (and even in the same model), and the cases of alternation are not always amenable to ordering.

Diphthong compression

In the general case, the compression process of ancient diphthongs can be represented as follows:

ai -> ē
aw -> ō
iy -> ī
uw -> ū

It is necessary to take into account that many ancient roots had w instead of y, and the models underwent various phonetic changes (in terms of vowels). In particular, most often the first root yod in diphthongs behaves like a vav.

At the roots ʔbd, ʔby, ʔhb, ʔxz, ʔkl, ʔmr, ʔpy in the imperfect 1 of the breed is always a and under the second root, and the prefix is ​​announced. In the forms of the 1st person singular, two alephas merge into one (and are announced). In the imperative and the infinitive act gzarot associated with laryngeal.

The roots with the third letter Yod or Aleph in the verbs change the preceding vocation when the endings begin after them, starting with a consonant. In the perfect: Yud in 1 and 2 rocks turns the previous vocalization into a full Hirik (ī), in the other rocks - into a full Cer (ē); the fallen Alef transforms the vocalization into Kamac (ā) in 1 breed, and in other breeds and with the median ē in the first breed - in the center (ē). In the imperfect and the imperative (the only ending to the consonant there is -nā), a segol is formed before the Yod and Aleph are dropped. All formed ceres are read as ē. Aleph and Yod are retained in writing, but not in pronunciation.

The processes of diphthong compression are not quite regular with the second weak letter in the root.

See also the section “Segolates and Quasi-Segolates”.

Assimilation

The letter Nun, as well as (in some roots) Yod and Lamed (at the root lqx), if they have a seam at rest and come after a letter with a small vow, can merge with the next letter (it gets a strong dagesh). This ghizra often does not work if the next letter is laryngeal (it cannot take dagesh).

In Binjana, the paul in the imperfect is the 1st root nun assimilated always, and in the imperative it happens differently: if the middle vocation сов, the nun drops out, if a does not (then the 2nd and 3rd root types 2a3 remain).

Letter Drop

Doubling

If the 2nd and 3rd letters coincide in the root, then if there is a resting seam (not fluttering) between them, they merge into one. When declining, a new form is used.

The letter Tav with the seam at rest merges with the letters Tav, Tet and Dalet, which in this case receive a strong dagesh.

3rd root Nun and Tav merge with the corresponding letters in the endings of the verb. At the root ntn, the final nun merges with both the tav and the nun of endings.

Gzarot table

In the table, ם means any letter, under ע - any of the guttural ones specified in the conditions.

Ashkenazi

The main differences between the Ashkenazic pronunciation of Hebrew and that adopted in Israel (a simplified version of Sephardic pronunciation) are as follows.

• The stress in Ashkenazic Hebrew during a conversation usually falls on the penultimate syllable, and when reading religious texts, in accordance with the Hebrew rules of grammar, usually the latter, and in Sephardic, the place of the ancient syllable remains (in most cases, the last syllable forms and in some categories of words - on the penultimate one. In the latter case, of course, the stress in the Ashkenazic and Sephardic variants is the same).
• In the Ashkenazic pronunciation, there is a difference in the pronunciation of the sound transmitted by the letter ת without the dagesh sign (“ Tav shafuya ”). In the pronunciation of Yemeni Jews, this letter is read as a slit sound similar to the English th in the word think . In Sephardic (and modern Israeli) pronunciation, the distinction between Tav Dgusha (that is, with the sign of Dagesh ) and Tav Rafuya was lost, and the letter ת is always read as T. In the Ashkenazi version, the slit pronunciation was preserved, albeit in a modified form - instead of the interdental sound θ , the sound С was pronounced.
• In ancient Hebrew there was a distinction of vowels in longitude, that is, vowels were long and short, according to the Vilna Gaon, long vowels are diphthongs . In modern Hebrew, there are no differences in vowel length, while the sound changes were different in the Sephardic and Ashkenazic versions. In the Sephardic version, the pronunciation of long vowels coincided with the pronunciation of short (except for the vowel Kamac , which has two options - long, “ Kamac gadol ”, and short, “ Kamac Katan ”, which was pronounced as “o”; in modern Hebrew, due to complexity the rules of the distinction of these vocalizations, “ kamaz katan ” is often replaced with the vocalization Holam ). Thus, in modern Hebrew “short a” and “long a” are pronounced the same way - as “a”. In the Ashkenazi variant, the long vowels A, O, and E changed their sound: long A was pronounced O (and then in southern dialects, for example, in Ukraine, it passed to U ); a long O passed into diphthong OH (and then in dialects in the territory of Lithuania and Belarus into diphthong AY ); long E went to diphong hey . The long sounds of U and Y in the Ashkenazic pronunciation coincided with the corresponding brief ones, that is, these two sounds are pronounced in the Ashkenazic version and in the Sephardic version in the same way.
• In addition, as a result of the aforementioned shift of emphasis, the vowel O, formed on the site of the original long A, underwent a reduction, and in words borrowed from Hebrew into Yiddish, was pronounced E (although in Hebrew texts, for example, when reading prayers, they continued to pronounce O ).

Examples

The word שבת ("Saturday") in Ashkenazic pronunciation sounds like "shabos" (Yiddish "shabes"), and in Sephardic - as "sabbath."

The word משפחה ("family") in the Ashkenazic pronunciation sounds like "mishpoho" (Yiddish "mishpokhe", "mishpuhe"), and in Sephardic - as "mishpaha".

The word בית-דין ("court") in Ashkenazic pronunciation sounds like "BASE-DIN", and in Sephardic - as "Beth-DIN."

The name משה in Ashkenazic pronunciation sounds like “Moshe”, and in Sephardic - as “ Moshe ”.

Sephardic

Eastern

The following features are highlighted in the pronunciation of the Jews of Arab countries. To these may be added certain features related to the local Arabic dialect.

• The emphasis is on the last syllable, wherever it may be in standard biblical Hebrew.
• א is pronounced laryngeal [[], except when used to denote a vowel.
• ב without Dagesh in some countries is pronounced as [b] (eg in Iraq), in others - as [v] (eg in Morocco).
• ג without dagesh is pronounced [ɣ] (Arabic غ).
• ד without Dagesh usually - [d], but sometimes it occurs [ð] (like Arabic ذ).
• ו in Iraq and Yemen [w], in other countries [v].
• ח pronounced [ħ] (Arabic).
• ט = [tˤ] = ط
• כ pronounced [x]
• ע = [ʕ] = ع
• צ = [sˤ] = ص
• ק is usually pronounced q \ ق, but sometimes it is found [k], [g], [ʕ].
• ר is usually trembling [r], but the Baghdad Jews pronounce it as uvular [ʀ], closer as Arabic غ.
• ת without Dagesh in some countries [t] (influenced by Sephardic and Arabic dialect pronunciation), in others (in Iraq, Yemen) - [θ] (ar.).
• The vowels are close to the Sephardic pronunciation, in particular, the ceres = [e:], Holam = [o:], great Kamac = [a:].

The pronunciation of Jews from non-Arab countries is different in some aspects. Thus, Persian Jews do not articulate Arabic [ħ] and [tˤ], and the great Kamac is pronounced as [ɒ] (long [ā] in Farsi ).

Yemen

Samaritan

Consonants

The consonants lost and developed during the development of Biblical Hebrew are highlighted.

The phonetic nature of some Hebrew consonants is controversial. The so-called "emphatic" consonants could be abruptive, pharyngalized or velarised. Some researchers believe that / s, z, sʼ / were pronounced as affricates / ts, dz, tsʼ / ^{[11] [12]} .

Initially, the Hebrew letters ח and ע designated two phonemes each, uvular and pharyngeal, the difference of which was not reflected in the spelling. By 200 BC. e. the sounds / χ / ח and / ʁ / ע have merged with their pharyngeal pairs / ħ / ח and / ʕ /. This is clearly seen in the Septuagint (for example, Isaac יצחק = Ἰσαάκ and Rachel רחל = Ῥαχήλ ) ^{[13] [14]} .

Similarly, the phoneme / ɬ /, denoted by the letter ש (also denotes / ʃ /), in late Biblical Hebrew merged with / s / (denoted by ס; in Samaritan pronunciation, sometimes / /). In transliteration, it is customary to denote this phoneme as / ś /. It was restored by comparison (it was in proto-Semitic and is still present in some South Arabian dialects), and also reflected in borrowings (Hebrew baśam, Greek balsamon "balm") ^{[9] [12] [15]} .

Allophonic spirantisation (/ b ɡ dkpt / in [v ɣ ð xf θ], the group “ runed kefet ”) developed in biblical Hebrew under the influence of Aramaic ^{[nb 1]} . Perhaps this happened after the loss of the original Aramaic / θ, ð / in the 7th century BC. e. ^[16] , and most likely after the disappearance of Hebrew / χ, ʁ /. It is known that this happened in the 2nd century AD. e. ^[17] After a certain period, this alternation became distinguishable in the middle and final position in the word (the functional difference was small), but in the initial position, the sounds “runaway kefet” continued to be allophones ^[18] . This is reflected in the Tiberias proclamation, where the words with the first letter of the “runaway kefet” have a spirant form after words ending in vowels. Moreover, Rabbi Saadia Gaon witnessed this alternation in Tiberias Aramaic at the beginning of the 10th century AD. e. ^[18]

The scrolls of the Dead Sea could alternate / ħ ʕ h ʔ / (for example, חמר ħmr in the Masoretic אָמַר / ʔɔˈmar / “he said”) ^[19] .

In Hebrew, the doubled consonants differed ^[20] . There are no double / wjz / fixed in Hexaplaces. In the Tiberian tradition, / ħ ʕ h ʔ r / do not double; historically, they first stopped doubling / ʔ r /, then / ʕ h /, later / ħ / (instead of doubling, these consonants change the previous vowels) ^[21] .

Vowels

The system of vowels of the Hebrew language has changed significantly over time. Below are the reconstructed initial vowels, vowels recorded in Hexaplach (in the second column, Secunda), the vowels of the three traditions of the vowing (Babylonian, Tiberias and Palestinian) and the vowels of the Samaritan tradition.

Phonetic shifts

Proto-Semitic language, a common ancestor of modern Semitic languages, according to traditional reconstruction had 29 consonants and 6 vowels-monophthongs (* / a aː i iː u uː /, long vowels are only in open syllables) and 2 diphthongs (* / aj aw / ) ^{[34] [35]} . The stress system is unknown, but it is usually described by an identical classical Latin or modern literary Arabic: if the penultimate syllable contains a consonant and a short vowel, then the stress falls on the third syllable from the end, otherwise the penultimate one.

Various, mostly morphological changes occurred during the transition from Proto-Semitic to Proto-Central Semitic. Phonology practically did not change, although the emphatic consonants could turn from abruptives into pharyngalized ones.

The morphological system of the Proto-Central Semitic language has changed significantly, especially in terms of the verb, and has become similar to the morphology of classical Arabic. The names inclined in three cases with the endings / -u, -a, -i /. In some situations (but never in a conjugate construct), the nasal sound, preserved in some languages ​​as nunation (/ -n /) or mimic (/ -m /), assumed names. The original meaning of this sound is unknown. In Arabic, it is an indicator of an indefinite state, and when using a definite article or another case, when a word becomes meaningfully definite, it falls out. In other languages, the final / -n / can be used to distinguish the absolute state from the conjugate state. The early biblical Hebrew (before 1500 BC) had an unknown function by name, for example, urušalemim (Jerusalem).

The Canaanite shift was the transition * / aː / b / oː / under unclear conditions ^{[nb 4]} . It occurred in the 14th century BC. Oe., it is reflected in the Amarn Archives (1365 BC. E.) ^{[36] [37]} .

After the Canaanite shift, the vowel system of the proto-Hebrew is reconstructed as * / a aː oː i iː u uː / (perhaps occasionally * / e: /) ^[22] . The emphasis was on the penultimate syllable, and modern stress on the last syllable is the result of the final vowels.

The final short vowels fell out in most words, because of this, long vowels began to be used in the closed stressed syllable. The fallout occurred in two stages: the markers of the verb forms began, then the markers of the cases. Between this there was a lengthening of vowels in the stressed open syllables (in names, not in verbs): * dabara ('the word' vin.p. ')> / dɔˈvɔr / (but * kataba (' he wrote ')> / kɔˈθav /).

Throwing off verb endings has worn out the differences between inclination forms, but also some generic forms. Vowel lengthening may be used to reflect differences. Thus, in suffixing conjugation, the first person singular * * -tu has been transferred to * -tī already in proto-Hebrew, by analogy with the possessive -ī (similarly, the word "I" * ana became * ʔanī). In the same way, the endings of the 2nd person * -ta -ti went to * -tā -tī for the male and female gender, or * -t for both. In monuments, the final long vowels could either be indicated or not indicated. In Hexaplahe both variants of endings are fixed (* t and * tā / tī). The spelling stabilized and preferred the spelling variant -t, while the speech preserved -tā for masculine and -t for feminine. The same process occurred in possessive * -ka ('your' m.) And * -ki ('your' f.) And personal * ʔanta, * ʔanti.

Short vowels * / aiu / tended to lengthen in the open pre-stressed and stressed syllable ^{[38] [nb 5]} . In the process of lengthening, the vowels of the upper ascent lowered. In Hexaplahe, an elongation of / aiu / a / a Г eː oː /; when the vowels remained short, they still dropped (/ aeo /) ^{[39] [nb 6] [nb 7]} .

Emphasis

• Dekel, Nurit. Colloquial Israeli Hebrew: A Corpus-based Survey . - De Gruyter, 2014. - ISBN 978-3-11-037725-5 .