Yemenite Hebrew

Yemenite Hebrew, also referred to as Temani Hebrew, is the pronunciation system for Hebrew traditionally used by Yemenite Jews. Yemenite Hebrew has been studied by language scholars, many of whom believe it retains older phonetic and grammatical features that have been lost elsewhere. Yemenite speakers of Hebrew are praised by language purists because of their use of grammatical features from classical Hebrew.
Some scholars believe that its phonology was heavily influenced by spoken Yemeni Arabic. Other scholars, including Yosef Qafih and Abraham Isaac Kook, hold the view that Yemenite Arabic did not influence Yemenite Hebrew, as this type of Arabic was also spoken by Yemenite Jews and is distinct from the liturgical and conversational Hebrew of the communities. Among other things, Qafih noted that the Yemenite Jews spoke Arabic with a distinct Jewish flavor, inclusive of pronouncing many Arabic words with vowels foreign to the Arabic language, e.g., the qamatz and tzere. He argues that the pronunciation of Yemenite Hebrew was not only uninfluenced by Arabic, but it influenced the pronunciation of Arabic by those Jews, despite the Jewish presence in Yemen for over a millennium.

History

Yemenite Hebrew may have been derived from, or influenced by, the Hebrew of the Talmudic academies in Babylonia: the oldest Yemenite manuscripts use the Babylonian vocalization, which is believed to antedate the Tiberian vocalization. As late as 937, Jacob Qirqisani wrote: "The biblical readings which are wide-spread in Yemen are in the Babylonian tradition." Indeed, in many respects, such as the assimilation of paṯaḥ and səġūl, the current Yemenite pronunciation fits the Babylonian notation better than the Tiberian. This is because in the Babylonian tradition of vocalization there is no distinct symbol for the səġūl. It does not follow, as claimed by some scholars, that the pronunciation of the two communities was identical, any more than the pronunciation of Sephardim and Ashkenazim is the same because both use the Tiberian symbols.
The following chart shows the seven vowel paradigms found in the Babylonian supralinear punctuation, which are reflected to this day by the Yemenite pronunciation of Biblical lections and liturgies, though they now use the Tiberian symbols. For example, there is no separate symbol for the Tiberian səġūl and the pataḥ and amongst Yemenites they have the same phonetic sound. In this connection, the Babylonian vowel signs remained in use in Yemen long after the Babylonian Biblical tradition had been abandoned, almost until our own time.

Vowels with ב
Tiberian equivalent	qamaṣ	paṯaḥ,	ṣerê	shewā mobile	ḥōlam	ḥiraq	šūraq, qubbūṣ
Tiberian niqqud		,					,
Value

Distinguishing features

The following chart shows the phonetic values of the Hebrew letters in the Yemenite Hebrew pronunciation tradition.

Letter

Value

Yemenites have preserved the sounds for each of the six double-sounding consonants: bəged-kəfet. The following are examples of their peculiar way of pronunciation of these and other letters:

gímel/ǧimal with the dāḡēš/dageš is pronounced. Thus, the verse is realized as, u'mi, ǧoi ǧaḏol .
gímel/gimal without dāḡēš/dageš is pronounced, like Arabic ġayn غ.
dāleṯ/dal without dāḡēš/dageš is pronounced, like Arabic ḏāl ذ. Thus, the word in Shema Yisrael is always pronounced aḥoḏ.
The pronunciation of tāv/taw without dāḡēš/dageš as ث . Thus, Sabbath day is pronounced in Yemenite Hebrew, Yom ha-Shabboth.
vāv/waw is pronounced .
Emphatic and guttural letters have nearly the same sounds and are produced from deep in the throat, as in Arabic.
ḥêṯ/ħet is a voiceless pharyngeal fricative, equivalent to Arabic ح.
ʻáyin/ʕajin is identical to Arabic ع, and is a voiced pharyngeal fricative.
tsadi is not a voiceless alveolar sibilant affricate "ts" among the Yemenites, but rather a deep-sounding "s".
qof is pronounced by the Yemenites as a voiced /g/, and is in keeping with their tradition that a different phonetic sound is given for gímel/''gimal.
resh is pronounced as an alveolar trill /r/, rather than the uvular trill, and is identical to Arabic ر '' and follows the conventions of old Hebrew.
The kaf sofit with a dagesh is pronounced as such, as in the rare example of the last word in Psalm 30.

Vowels

Qāmaṣ gāḏôl/Qamac qadol is pronounced, as in Ashkenazi Hebrew and Tiberian Hebrew. The Yemenite pronunciation for Qamats gadol and Qamats qatan is identical.
There is no distinction between the vowels paṯaḥ/pataħ and səḡôl/segol all being pronounced, like the Arabic fatḥa. A šəwâ nāʻ/šwa naʕ, however, is identical to a חטף פתח and חטף סגול.
Final hê/hej with mappîq/mefiq has an aspirated sound, generally stronger sounding than the regular hê/hej. Aleph with a dagesh, a rare occurrence, is pronounced with a glottal stop, e.g., the word וַיָּבִיאּוּ in . Conversely, some words in Hebrew which are written with the final hê ending are realized by a secondary glottal stop and so are abruptly cut short, as to hold one's breath.
A semivocalic sound is heard before paṯaḥ gānûḇ/pataħ ganuv : thus ruaħ sounds like rúwwaḥ and siaħ sounds like síyyaḥ.

Yemenite pronunciation is not uniform, and Morag has distinguished five sub-dialects, the best known being probably Sana'ani, originally spoken by Jews in and around Sana'a. Roughly, the points of difference are as follows:

In some dialects, ḥōlem/ħolam is pronounced, but in others, it is pronounced like ṣêrệ/cerej.
Some dialects do not differentiate between bêṯ/bet with dāḡēš/dageš and without it. That occurs most of Mizrahi Hebrew.
Sana'ani Hebrew primarily places stress on the penultimate syllable, as in Ashkenazi Hebrew.

''Qamats Gadol'' and ''Qamats Qatan''

Yemenite reading practices continue the orthographic conventions of the early grammarians, such as Abraham ibn Ezra and Aaron Ben-Asher. One basic rule of grammar states that every word with a long vowel sound, that is, one of either five vowel sounds whose mnemonics are "pīt'ūḥe' ḥöthom", whenever there is written beside one of these long vowel sounds a meteg and is denoted by a small vertical line below the word, it indicates that the vowel must be drawn out with a prolonged sound. For example, ōōōōōō, instead of ō,. In the Sephardic tradition, however, the practice is different altogether, and they will also alter the phonetic sound of the short vowel qamaṣ qattön whenever the vowel appears alongside a meteg, reading it as the long vowel qamaṣ gadöl, giving to it the sound of "a", as in car, instead of "ōōōōō." Thus, for the verse in, the Sephardic Jews will pronounce the word כָּל as "kal", instead of kol ʕaṣmotai as pronounced by both Yemenite and Ashkenazi Jewish communities.
The meteg, or ga’ayah, has actually two functions: It extends the sound of the vowel; It makes any šewa that is written immediately after the vowel a mobile šewa, meaning, the šewa itself becomes [ə|]. For example: = ʔö m^ǝ rim, = šö m^ǝ rim, = sī s^ǝ ra, = šū v^ǝ kha, and = tū v^ǝ kha. Examples with meteg/ga’ayah: = šoː m^ǝ ro, = ye r^ǝ du.
The Qamats qatan is realized as the non-extended "o"-sound in the first qamats in the word, חָכְמָה ⇒ ḥokhma.
The Yemenite qamaṣ is represented in the transliterated texts by the diaphoneme. The vowel quality is the same, whether for a long or short vowel, but the long vowel sound is always prolonged.

''Holam'' and ''sere''

A distinct feature of Yemenite Hebrew is that there is some degree of approximation between the ḥōlam and the ṣêrệ. To the untrained ear, they may sound as the same phoneme, but Yemenite grammarians will point out the difference. The feature varies by dialect:

In the standard, provincial pronunciation that is used by most Yemenite Jews, holam is pronounced as. For example, the word "shalom", is pronounced sholøm.
In some provincial dialects, in particular that of Aden, holam becomes a long e and is indeed indistinguishable from sere, and some early manuscripts sometimes confuse or interchange the symbols for the two sounds.

Some see the assimilation of the two vowels as a local variant within the wider Babylonian family, which the Yemenites happened to follow.

Strict application of Mobile Shewā

Rabbi Abraham Isaac Kook and Rabbi Jacob Saphir have praised the Yemenites in their correct pronunciation of Hebrew. They still read the biblical lections and liturgies according to what is prescribed for Hebrew grammar and are meticulous to pronounce the mobile šĕwā in each of its changing forms. While most other communities also adhere to the rule of mobile šĕwā whenever two šĕwās are written one after the other, as in, most have forgotten its other usages.
Aharon Ben-Asher, in his treatise on the proper usage of Hebrew vowels and trope symbols, writes on the šĕwā: " the servant of all the letters in the entire Scriptures, whether at the beginning of the word, or in the middle of the word, or at the end of the word; whether what is pronounced by the tongue or not pronounced, for it has many ways… However, if it is joined with one of four letters, א ח ה ע, its manner will be like the manner of the vowel of the second letter in that word, such as: = b^öhonoth; = t^e’ehavu; = l^eḥeləkhah; = r^eʻeloyoh."
On the mobile šĕwā and its usage amongst Yemenite Jews, Israeli grammarian Shelomo Morag wrote: "The pronunciation of the šĕwā mobile preceding א, ה, ח, ע, or ר in the Yemenite tradition is realized in accordance with the vowel following the guttural; quantitatively, however, this is an ultra-short vowel. For example, a word such as is pronounced w^uḥuṭ. A šĕwā preceding a yōḏ is pronounced as an ultra-short ḥīreq: the word is pronounced bⁱyōm. This is the way the šĕwā is known to have been pronounced in the Tiberian tradition."
Other examples of words of the mobile šĕwā in the same word taking the phonetic sound of the vowel assigned to the adjacent guttural letter or of a mobile šĕwā before the letter yod taking the phonetic sound of the yod, can be seen in the following:

= w^eheshiv
= bam^oʻoroh
= b^eʻevar
= wⁱ’im
= mizbⁱḥī
= w^ohoyoh

*
lⁱyöm – l^ohödöth – l^ahaǧīd
The above rule applies only to when one of the four guttural letters, or a yod or a resh follows the mobile šĕwā, but it does not apply to the other letters; then, the mobile šĕwā is always read as a short-sounding pataḥ.

Distinctive pronunciations preserved

Geographically isolated for centuries, the Yemenite Jews constituted a peculiar phenomenon within Diaspora Jewry. In their isolation, they preserved specific traditions of both Hebrew and Aramaic. The traditions, transmitted from generation to generation through the teaching and reciting of the Bible, post-biblical Hebrew literature, the Aramaic Targums of the Bible, and the Babylonian Talmud, are still alive. They are manifest in the traditional manner of reading Hebrew that is practised by most members of the community. The Yemenite reading traditions of the Bible are now based on the Tiberian text and vocalization, as proofread by the masorete, Aaron ben Asher, with the one exception that the vowel sǝġūl is pronounced as a pataḥ, since the sǝġūl did not exist in the Babylonian orthographic tradition to which the Jews of Yemen had previously been accustomed. In what concerns Biblical orthography, with the one exception of the sǝgūl, the Yemenite Jewish community does not differ from any other Jewish community.
Although the vast majority of post-Biblical Hebrew and Aramaic words are pronounced the same way or nearly the same way by all of Israel's diverse ethnic groups, including the Jews of Yemen, there are still other words whose phonemic system differs greatly from the way it is used in Modern Hebrew, the sense here being the tradition of vocalization or diction of selective Hebrew words found in the Mishnah and Midrashic literature, or of Aramaic words found in the Talmud, and which tradition has been meticulously preserved by the Jews of Yemen. Two of the more recognized Yemenite pronunciations are for the words rtl=yes and rtl=yes, the first pronounced as Ribbi, instead of Rabbi, and the second pronounced guvra, instead of gavra. In the first case, archaeologist Benjamin Mazar was the first to discover its linguistic usage in the funerary epigrams of the 3rd and 4th-century CE, during excavations at the catacombs in Beit She'arim. Nahman Avigad, speaking of the same, wrote: "Of special interest is the title Rabbi and its Greek transliteration. In the inscriptions of Beth She'arim found in the former seasons rtl=yes and rtl=yes are usual, and only once do we find rtl=yes, which has been regarded as a defective form of rtl=yes, for in Greek we generally find the form. The transliteration found here shows that the title was pronounced in Palestine in different ways, sometimes Rabbi, sometimes Ribbi and occasionally even Rebbi." In the latter case, the Jerusalem Talmud occasionally brings down the word rtl=yes in plene scriptum, rtl=yes, showing that its pronunciation was the same as that in use by the Yemenites. Some have raised the proposition that the Yemenite linguistic tradition dates back to the Amoraim.
R. Yehudai Gaon, in his Halakhot Pesukot, uses yod as the mater lectionis to show the vowel hiriq, after the qoph in Qiryat Shema. The editor of the critical edition, A. Israel, who places its composition in Babylonia, notes that "linguists would take an interest" in Yehudai Gaon's variant spellings of words, where especially the matres lectionis is used in place of vowels, "represented either by a plene alef, waw, and yod." The use of the matres lectionis in place of the vowel hiriq in the construct case of the words קִרְיַת שְׁמַע reflects apparently the Babylonian tradition of pronunciation, and, today, the same tradition is mirrored in the Yemenite pronunciation of Qiryat shemaʻ.
The following diagrams show a few of the more conspicuous differences in the Yemenite tradition of vocalization and which Israeli linguist, Shelomo Morag, believes reflects an ancient form of vocalizing the texts and was once known and used by all Hebrew-speakers.

Notes on transliteration: In the Yemenite Jewish tradition, the vowel qamaṣ, represents. The Hebrew character Tau, without a dot of accentuation, represents. The Hebrew character Gimal, with a dot of accentuation, represents. The Hebrew word rtl=yes, is written in Yemenite Jewish tradition with a vowel qamaṣ beneath the, but since it is followed by the letters rtl=yes it represents. The vowel ḥolam in the Yemenite dialect is transcribed here with, and represents a front rounded vowel. Another peculiarity with the Yemenite dialect is that the vast majority of Yemenite Jews will replace, used here in transliteration of texts, with the phonetic sound of.

In the Yemenite tradition, the plural endings on the words rtl=yes, rtl=yes, rtl=yes, rtl=yes, rtl=yes and rtl=yes, all differ from the way they are vocalized in Modern Hebrew. In Modern Hebrew, these words are marked with a shuraq, as follows: rtl=yes – rtl=yes – rtl=yes – rtl=yes – rtl=yes – rtl=yes. Although the word in Daniel 8:22 is vocalized malkhuyoth, as it is in Modern Hebrew, Shelomo Morag thinks that the Yemenite tradition reflects a phonological phenomenon known as dissimilation, whereby similar consonants or vowels in a word become less similar. Others explain the discrepancy as being in accordance with a general rule of practice, prevalent in the 2nd century CE, where the Hebrew in rabbinic literature was distinguished from that of Biblical Hebrew, and put into an entire class and category of its own, with its own rules of vocalization.
The Hebrew noun rtl=yes, in the upper left column, is a word meaning "slice/piece", or rtl=yes in the construct state. The noun is of the same metre as rtl=yes, a word meaning "peel," or the "rind" of a fruit. Both the kaph and pe in these nouns are with a dagesh. However, the same roots applied to different meters, serving as gerunds, as in "slicing/cutting" and "peeling", the words would respectively be rtl=yes and rtl=yes, without a dagesh in the Hebrew characters Kaph and Pe, such as when the verb is used with the preposition "after": e.g. "after peeling the apple" = rtl=yes, or "after cutting the meat" = rtl=yes.

In the Talmud, the Sages of Israel had a practice to read words derived from the Scriptures in their own given way, while the same words derived from the Talmud or in other exegetical literature in a different way: "When Isse the son of Hinei went up, he found Rabbi Yoḥanan teaching to the creations, saying, raḥelim, etc. He said to him, 'Teach it, raḥeloth!' He replied, ' as it is written : Ewes, two-hundred.' He answered him, 'The language of the Torah is by itself, and the language employed by the Sages is by itself!'".
This passage from the Talmud is often quoted by grammarians of Yemenite origin to explain certain "discrepancies" found in vocalization of words where a comparable source can be found in the Hebrew Bible, such as the Yemenite tradition in rabbinic literature to say, rather than – although the latter rendering appears in Scripture, or to say, with ḥīraq, rather than,, with ṣerê, although it too appears in Scripture, or to say, rather than as the word "blessing" in the construct state which appears in the Scriptures, e.g. birkath Avraham, with kaph dagesh. Others, however, say that these anomalies reflect a tradition that antedates the Tiberian Masoretic texts.
Along these same lines, the Masoretic Text of the Hebrew Bible renders the words, in II Chronicles 26:6, and, in Nehemiah 7:37; 11:35, as yävnɛ and lōð, respectively. However, in their demotic-forms, the Yemenites will pronounce these words as and = yovnei and lūd, respectively. The use of the phoneme "ṣerê", represented by the two dots "◌ֵ", instead of "pataḥ-səġūl" for the word "Yavneh" may have been influenced by the Palestinian dialect spoken in the Land of Israel in the 1st-century CE.

In Yemenite tradition, many words in both Biblical and Mishnaic Hebrew which are written with the final hê ending are realized by a secondary glottal stop, meaning, they are abruptly cut short, as when one holds his breath. Shelomo Morag who treats upon this peculiarity in the Yemenite tradition of vocalization brings down two examples from the Book of Isaiah, although by no means exclusive, where he shows the transliteration for the words rtl=yes in Isaiah 1:27 and rtl=yes in Isaiah 2:5, and both of which represent, as in tippoːdä and wǝnel^ăχoː respectively. The word rtl=yes in the upper-middle column is pronounced in the same way, e.g..

Excursus: The preposition is unique in the Yemenite Jewish tradition. The Hebrew preposition is always written with the noun, joined as one word, and the lamed is always accentuated with a dagesh. For example, if the noun, would normally have been written with the definite article, as in, and the noun was to show possession, as in the sentence: "the palace of the king," the definite article "the" is dropped, but the same vowel pataḥ of the definite article is carried over to the lamed, as in rtl=yes, instead of של המלך. The vowel on the lamed will sometimes differ, depending on what noun comes after the preposition. For example, the definite article "the" in Hebrew nouns which begin with aleph or resh and sometimes ayin, such as in rtl=yes and in rtl=yes, or in rtl=yes, is written with the vowel qamaṣ – in which case, the vowel qamaṣ is carried over to the lamed, as in rtl=yes and in rtl=yes and in rtl=yes. Another general rule is that whenever a possessive noun is written without the definite article "the", as in the words, "a king's sceptre," or "the sceptre of a king", the lamed in the preposition is written with the vowel shǝwa, as in rtl=yes, and as in, "if it belongs to Israel" ⇒ rtl=yes. Whenever the noun begins with a shǝwa, as in the proper noun Solomon and one wanted to show possession, the lamed in the preposition is written with a ḥiraq, as in : rtl=yes ⇒ "Solomon's bed", or as in rtl=yes ⇒ "the punishment rtl=yes the wicked", or in rtl=yes ⇒ "a bundle rtl=yes heave-offering."
Another rule of practice in Hebrew grammar is that two shǝwas rtl=yes are never written one after the other at the beginning of any word; neither can two ḥaṭaf pataḥs rtl=yes or two ḥaṭaf sǝġūls rtl=yes be written at the beginning of a word one after the other. The practical implication arising from this rule is that when there is a noun beginning with a ḥaṭaf pataḥ, as in the word, rtl=yes ⇒ "her companion", and one wishes to add thereto the preposition "to" – as in, "to her companion" ⇒ rtl=yes, the lamed is written with the vowel pataḥ, instead of a shǝwa, seeing that the shǝwa at the beginning of a word and the ḥaṭaf pataḥ, as well as the ḥaṭaf sǝġūl, are all actually one and the same vowel, and it is as though he had written two shǝwas one after the other. Likewise, in the possessive case, "belonging to her companion" ⇒ rtl=yes, the lamed in the preposition rtl=yes is written with the vowel pataḥ.

Hebrew vernacular

The Leiden MS. of the Jerusalem Talmud is important in that it preserves some earlier variants to textual readings of that Talmud, such as in Tractate Pesaḥim 10:3, which brings down the old Palestinian-Hebrew word for charoseth, viz. dūkeh, instead of rūbeh/rabah, saying with a play on words: "The members of Isse's household would say in the name of Isse: Why is it called dūkeh? It is because she pounds with him." The Hebrew word for "pound" is dakh, which rules out the spelling of " rabah ", as found in the printed editions. Today, the Jews of Yemen, in their vernacular of Hebrew, still call the charoseth by the name dūkeh.
Other quintessential Hebrew words which have been preserved by the Jews of Yemen is their manner of calling a receipt of purchase by the name, roʔoːyoː, rather than the word "qabbalah" that is now used in Modern Hebrew. The weekly biblical lection read on Sabbath days is called by the name seder, since the word parashah has a completely different meaning, denoting a Bible Codex containing the first Five Books of Moses.
Charity; alms, so-called in Yemenite Jewish parlance, was usually in the form of bread, collected in baskets each Friday before the Sabbath by those appointed over this task for distribution among the needy, without them being brought to shame. The same word is often used throughout the Jerusalem Talmud, as well as in Midrashic literature, to signify what is given out to the poor and needy. Today, in Modern Hebrew, the word is seldom used to imply charity, replaced now by the word, ts’dakah. In contrast, the word rtl=yes amongst Jews in Sana’a was a tax levied upon Jewish householders, particularly those whose professions were butchers, and which tax consisted of hides and suet from butchered animals, and which things were sold on a daily basis by the Treasurer, and the money accruing from the sale committed to the public fund for the Jewish poor of the city, which money was distributed to the city's poor twice a year; once on Passover, and once on Sukkot. The fund itself was known by the name toːḏer, lit. "the constant ."
Although Jews in Yemen widely made-use of the South-Arabic word mukhwāṭ for the "metal pointer" used in pointing at the letters of sacred writ, they also knew the old Hebrew word for the same, which they called makhtev. The following story is related about this instrument in Midrash Rabba: "Rabban Gamliel says: ‘Five-hundred schools were in Beter, while the smallest of them wasn’t less than three-hundred children. They used to say, ‘If the enemy should ever come upon us, with these metal pointers we’ll go out against them and stab them!’..."
In other peculiar words of interest, they made use of the word, shilṭön, for "governor" or "king," instead of "government," the latter word now being the more common usage in Modern Hebrew; kothev, for "scrivener", or copyist of religious texts, instead of the word "sofer" ; ṣibbūr, for "a quorum of at least ten adult males," a word used in Yemen instead of the Modern Hebrew word, minyan; ḥefeṣ, a noun meaning "desirable thing," was used by them to describe any "book", although now in Modern Hebrew it means "object"; fiqfūq had the connotation of "shock," "violent agitation," or "shaking-up," although today, in Modern Hebrew, it has the meaning of "doubt" or "skepticism"; the word, harpathqe, was used to describe "great hardships," although in Modern Hebrew the word has come to mean "adventures." The word fazmūn, any happy liturgical poem, such as those sung on Simhat Torah, differs from today's Modern Hebrew word, pizmon, meaning, a "chorus" to a song. Another peculiar aspect of Yemenite Hebrew is what concerns denominative verbs. One of the nouns used for bread is himmuṣ, derived from the blessing that is said whenever breaking bread, rtl=yes = He that brings forth . Whenever they wanted to say its imperative form, "break bread!", they made use of the denominative verb hammeṣ!. Similarly, the noun for the Third Sabbath meal was qiyyūm, literally meaning "observance," in which they made use of the denominative verb, tǝqayyem = Will you eat with us ?, or, rtl=yes = Let us eat, or, qiyam = He ate.