Gesenius' Hebrew Grammar/9. Character of the several Vowels
9a Numerous as are the vowel signs in Hebrew writing, they are yet not fully adequate to express all the various modifications of the vowel sounds, especially with respect to length and shortness. To understand this better a short explanation of the character and value of the several vowels is required, especially in regard to their length and shortness as well as to their changeableness (§§ 25, 27).
I. First Class. A-sound.
1. Qameṣ (־ָ), when it represents a long a, is, by nature and origin, of two kinds:—
(1) The essentially long â (in Arabic regularly written ־ָא), which is not readily shortened and never wholly dropped (§25c), e.g. כְּתָב kethâbh (writing); very seldom with a following א, as רָאשׁ 2 S 121, 4 (see the examples in §72p).
9b The writing of קָאם Ho 1014 for קָם would only be justifiable, if the ā of this form were to be explained as a contraction of ăă cf. however §72a; דָּאג Neh 1316 for דָּגּ (dāg) is certainly incorrect.—The rarity of the â in Hebrew arises from the fact that it has for the most part become an obtuse ô; see below, q.
9c (2) ā, lengthened only by position (i.e. tone-long or at all events lengthened under the influence of the tone, according to the laws for the formation of syllables, §27e–h), either in the tone-syllable itself (or in the secondary tone-syllable indicated by Mèthĕg, see below), or just before or after it. This sound is invariably lengthened from an original ă, and is found in open syllables, i.e. syllables ending in a vowel (§26b), e.g. לְךָ, קָטַל, יָקוּם, אָסִיר (Arab. lăkă, qătălă, yăqûmŭ, ʾăsîrŭ), as well as in closed syllables, i.e. those ending in a consonant, as יָד, כּוֹכָב (vulgar Arab. yăd, kaukăb). In a closed syllable, however, it can only stand when this has the tone, דָּבָ֫ר, עוֹלָ֫ם; whereas in an open syllable it is especially frequent before the tone, e.g. דָּבָ֫ר, זָקֵ֫ן, לָכֶ֫ם. Where the tone is moved forward or weakened (as happens most commonly in what is called the construct state of nouns, cf. §89a) the original short ă (Pathaḥ) is retained in a closed syllable, while in an open syllable it becomes Šewa (§27i): חָכָם, constr. state חֲכָס (ḥakhăm); דָּבָר, דְּבַר (debhăr); קָטַל, קְטָלָם. For examples of the retention, in the secondary tone-syllable, of ā lengthened from ă, see §93xx.
9d In some terminations of the verb (תָּ in the 2nd sing. masc. perf., ןָ in the 2nd pl. fem. of the imperat., as well as in the 3rd and 2nd pl. fem. of the imperf.), in אַתָּ thou (masc.) and in the suffixes ךָ and הָ, the final ā can stand even without a vowel letter. A ה is, however, in these cases (except with הָ) frequently added as a vowel letter.
On ־ָ for ŏ see below, f.
9e 2. Pathaḥ, or short ă, stands in Hebrew almost exclusively in a closed syllable with or without the tone (קָטַ֫ל, קְטַלְתֶּ֫ם). In places where it now appears to stand in an open syllable the syllable was originally closed, and a helping vowel (ă, ĭ) has been inserted after the second radical merely to make the pronunciation easier, e.g. נַ֫חַל (ground-form naḥl), בַּ֫יִת (Arab. bait), see §28d, and with regard to two cases of a different kind, §25g, g. Otherwise ă in an open syllable has almost without exception passed into ā (־ָ), see above, c.
9f 3. Segôl (ĕ, è [ǟ]) by origin belongs sometimes to the second, but most frequently to the first vowel class (§27o, o, u). It belongs to the first class when it is a modification of a (as the Germ. Bad, pl. Bäder; Eng. man, pl. men), either in a toneless syllable, e.g. יֶדְכֶם (for yadkhèm), or with the tone, e.g. אֶ֫רֶץ from ʾarṣ, קֶ֫רֶן Arab. qărn, קֶ֫מַח Arab. qămḥ. This Segôl is often retained even in the strongest tone-syllable, at the end of a sentence or of an important clause (in pause), as מֶ֑לֶךְ, צֶֽדֶק (mā̈́lä̆kh, sā̈́dä̆q). As a rule, however, in such cases the Pathaḥ which underlies the Segôl is lengthened into Qameṣ, e.g. קָ֑מַח, קָֽרֶן. A Segôl apparently lengthened from Šewa, but in reality traceable to an original ă, stands in pausal forms, as פֶּֽרִי (ground-form păry), יֶֽהִי (yăhy), &c. On the cases where a י (originally consonantal) follows this Segôl, see §75f, and §91k.
9g 4. The long î is frequently even in the consonantal writing indicated by י (a fully written Ḥireq ־ִי); but a naturally long î can be also written defectively (§8i), e.g. צַדִּיק (righteous), plur. צַדִּקִים ṣaddîqîm; יִירָא. (he fears), plur. יִרְֽאוּ. Whether a defectively written Ḥireq is long may be best known from the origin of the form; often also from the nature of the syllable (§ 26), or as in יִֽרְאוּ. from the Metheg attached to it (§16f).
9h 5. The short Ḥireq (always written defectively) is especially frequent in sharpened syllables (קִטֵּל, אִמִּי) and in toneless closed syllables (מִזְמוֹר psalm); cf. however וַיִּשְׁבְּ in a closed tone-syllable, and even וַיִּ֫פֶן, with a helping Segôl, for wayyiphn. It has arisen very frequently by attenuation from ă, as in דִּבְרֵי from original dăbărê, צִדְקִי (ground-form ṣădq), or else it is the original ĭ, which in the tone-syllable had become ē, as in אֹֽיִבְךָ (thy enemy) from אֹיֵב (ground-form ’âyĭb)
It is sometimes a simple helping vowel, as in בַּ֫יִת, §28e.
The earlier grammarians call every Ḥireq written fully, Ḥireq magnum; every one written defectively, Ḥireq parvum,—a misleading distinction, so far as quantity is concerned.
9i 6. The longest ê ־ֵי (more rarely defective ־ֵ, e.g. עֵנֵי for עֵינֵי Is 38; at the end of a word also ־ה) is as a rule contracted from ־ַי ay (ai), §7a, e.g. הֵיכָל (palace), Arab. and Syriac haikal.
9k 7. The Ṣere without Yôdh mostly represents the tone-long ē, which, like the tone-long ā (see c), is very rarely retained except in and before the tone-syllable, and is always lengthened from an original ĭ. It stands in an open syllable with or before the tone, e.g. סֵ֫פֶר (ground-form sĭphr) book, שֵׁנָ֫ה (Arab. sĭnăt) sleep, or with Metheg (see §16d, f) in the secondary tone-syllable, e.g. שְׁאֵֽלָתִי my request, נֵֽלְכָה let us go. On the other hand in a closed syllable it is almost always with the tone, as בֵּן son, אִלֵּם dumb.
9l Exceptions: (a) ē is sometimes retained in a toneless closed syllable, in monosyllabic words before Maqqeph, e.g. עֵֽץ־ Nu 3518, as well as in the examples of nāsôg ’āḥôr mentioned in §29f (on the quantity cf. §8b 3 end); (b) in a toneless open final syllable, Ṣere likewise occurs in examples of the nāsôg ’āḥôr, as יֵ֫צֵא Ex 1629; cf. Ju 939.
9m 8. The Segôl of the I(E)-class is most frequently an ĕ modified from original ĭ, either replacing a tone-long ē which has lost the tone, e.g. תֶּן־ from תֵּן (give), יֹֽצֶרְךָ (thy creator) from יֹצֵר, or in the case discussed in §93o, הֶלְקִי, עֶזְרִי from the ground-forms ḥilq, ‛izr; cf. also §64f. Segôl appears as a simple helping-vowel in cases such as סֵ֫פֶר for siphr, יִ֫גֶל for yigl (§28e).
9n 9. For the U-sound there is—
(1) the long û, either (a) written fully, וּ Šureq, e.g. גְּבוּל (boundary), or (b) defectively written ־ֻ Qibbûṣ גְּבֻלוֹ, יְמֻתוּן;
(2) the short ŭ, mostly represented by Qibbûṣ, in a toneless closed syllable and especially common in a sharpened syllable, in e.g. שֻׁלְחָן (table), סֻכָּה (booth).
For this u the LXX write o, e.g. עֲדֻלָּם Ὀδολλάμ, from which, however, it only follows, that this ŭ was pronounced somewhat indistinctly. The LXX also express the sharp Ḥireq by ε, e.g. אִמֵּר=Ἐμμήρ. The pronunciation of the Qibbûṣ like the German ü, which was formerly common, is incorrect, although the occasional pronunciation of the U-sounds as ü in the time of the punctators is attested, at least as regards Palestine; cf. the Turkish bülbül for the Persian bulbul, and the pronunciation of the Arabic dunyā in Syria as dünyā.
9p 10. The O-sound bears the same relation to U as the E does to I in the second class. It has four varieties:—
(1) The ô which is contracted from aw (=au), §7a, and accordingly is mostly written fully; וֹ (Holem plenum), e.g. שׁוֹט (a whip), Arab. sauṭ, עוֹלָה (iniquity) from עַוְלָה. More rarely defectively, as שֹֽׁרְךָ (thine ox) from שׁוֹר Arab. ṯaur.
9q (2) The long ô which arose in Hebrew at an early period, by a general process of obscuring, out of an original â, while the latter has been retained in Arabic and Aramaic. It is usually written fully in the tone-syllable, defectively in the toneless, e.g. קֹטֵל Arab. qâtĭl. Aram. qâṭēl, אֱלוֹהַּ Arab. ’ĭlâh, Aram. ’ĕlâh, plur. אֱלֹהִים; שׁוֹק (leg), Arab. sâq; גִּבּוֹר (hero), Arab. găbbâr; חוֹתָם (seal), Arab. ḫâtăm; רִמּוֹן (pomegranate), Arab. rŭmmân; שִׁלְטוֹן (dominion), Aram. שֻׁלְטָן and שָׁלְטָן Arab. sŭlṭân; שָׁלוֹם (peace), Aram. שְׁלָם, Arab. sălâm. Sometimes the form in â also occurs side by side with that in ô as שִׁרְיָן and שִׁרְיוֹן (coat of mail; see however §29u). Cf. also §68b.
9r (3) The tone-long ō which is lengthened from an original ŭ, or from an ŏ arising from ŭ, by the tone, or in general according to the laws for the formation of syllables. It occurs not only in the tone-syllable, but also in an open syllable before the tone, e.g. קֹדֶשׁ (ground-form qŭdš) sanctuary; בֹּרַךְ for burrakh, יִלְקֹטוּן ψ 10428, as well as (with Metheg) in the secondary tone-syllable; אֹֽהָלִים, פֹּֽעֲלוֹ. But the original ŏ (ŭ) is retained in a toneless closed syllable, whereas in a toneless open syllable it is weakened to Šeŵa. Cf. כֹּל all, but כָּל־ (kŏl), כֻּלָּם (kŭllām); יִקְטֹל, יִקְטָלְךָ and יִקְטְלוּ, where original ŭ is weakened to Šeŵa: yiqɩ̣̇elû, Arab. yaqtŭlû. This tone-long ō is only as an exception written fully.
9s (4) ־ָ Qameṣ-ḥaṭuph represents ŏ (properly å̆, cf. §8a, note 2) modified from ŭ and is therefore classed here. It stands in the same relation to Ḥolem as the Segôl of the second class to Sere, כָּל־ kŏl, וַיָּ֫קָם wayyāqŏm. On the distinction between this and Qameṣ, see below, u.
9t 11. The following table gives a summary of the gradation of the three vowel-classes according to the quantity of the vowels:—
|First Class: A.||Second Class: I and E.||Third Class: U and O.|
|־ָ original â (Arabic ־ָא).||־ֵי ê, from original ay (ai).
־ִי or ־ִ long î.
|וֹ ô, from original aw (au).
וֹ or ־ֹ ô obscured from â.
וֹ or ־ֻ û
|־ָ tone-long ā (from original ă) chiefly in the tone-syllable but also just before it.||־ֵ tone-long ē (from ĭ generally in the tone-syllable but also just before it.||־ֹ tone-long ō (from original ŭ in the tone-syllable, otherwise in an open syllable.|
|־ֶ (as a modification of ă) sometimes a tone-long è, sometime ĕ
־ַ short ă
־ִ ĭ attenuated from ă; see h.
Utmost weakening to ־ֲ a, ־ֱ ĕ, ־ְ e.
־ִ short ĭ
Utmost weakening to ־ֲ a, ־ֱ ĕ, or ־ְ e.
|־ָ ŏ, modified from ŭ
־ֻ short ŭ, especially in a sharpened syllable.
Utmost weakening to ־ֲ a, ־ֱ ĕ, ־ֳ o, or ־ְ e.
9u Rem. On the distinction between Qameṣ and Qameṣ-ḥaṭuph.
According to §8a, long ā or ā̊ (Qameṣ) and short ŏ or å̆ (Qameṣ-ḥaṭuph) are in manuscripts and printed texts generally expressed by the same sign ( ָ ), e.g. קָם qām, כָּל־ kŏl. The beginner who does not yet know the grammatical origin of the words in question (which is of course the surest guide), may depend meanwhile on the following principal rules:—
(a) When Šewâ follows as a syllable-divider, as in חָכְמָ֫ה ḥŏkh-mā́ (wisdom), אָכְלָ֫ה ’ŏkh-lā́ (food). With Metheg ־ָ is ā (å̄) and according to the usual view stands in an open syllable with a following Šewâ mobile, e.g. אָֽכְלָה ’ā-khelā́ (she ate); but cf. §16i.
(b) When a closed syllable is formed by Dageš forte, e.g. הָנֵּ֫נִי ḥŏnnēnî (have mercy upon me); but בָּֽתִּ֫ים (with Metheg, §16f ζ) bâttîm.
(c) When the syllable in question loses the tone on account of a following Maqqēph (§16a), e.g. כָּל־הָֽאָדָם kŏl-hā-’ādā́m (all men).
In ψ 3510 and Pr 197 Maqqēph with כָּל is replaced by a conjunctive accent (Merekha); so by Darga, Ju 195 with סְעָד, and Ez 378 with וַיִּקְרָם (so Baer after Qimḥi; ed. Mant., Ginsburg, Kittel ויקרַם).
(d) In a closed final syllable without the tone, e.g. וַיָּ֫קָם wayyā́qŏm (and he stood up).—In the cases where â or ā in the final syllable has become toneless through Maqqēph (§16a) and yet remains, e.g. כְּתָֽב־הַדָּת Est 48, שָֽׁת־לִי Gn 425, it has a Metheg in correct manuscripts and printed texts.
In cases like הָ֫לְאָה, לָ֫מָּה lā́mmā, the tone shows that ־ָ is to be read as ā.
9v 2. The cases in which ־ָ appears to stand in an open syllable and yet is to be read as ŏ require special consideration. This is the case, (a) when Ḥaṭeph-Qameṣ follows, e.g. פָּֽעֳלוֹ his work, or simple vocal Šewâ, e.g. דָּֽרְבָן ox goad; בְּעָֽבְרוֹ Jo 47; שָֽׁמְרָה (so ed. Mant., Ginsb.) preserve ψ 862, cf. 16:1 and the cases mentioned in §48i, n., and §61f, n.; other examples are Ob 111, Ju 1415); Ḥaṭeph-Pathaḥ follows in לִמְשָֽׁחֲךָ (so Ginsburg; Baer לִמְשָֽׁחֳךָ) 1 S 151, לַֽהֲרָֽגֲךָ 24:11, and יִֽפְגָֽשֲׁךָ (so Baer, Gn 3218, others יִפְגָּֽשְׁךָ); (b) before another Qameṣ-ḥaṭuph, e.g. פָּֽעָלְךָ֫ thy work; on אָֽרָה־לִּי and קָֽבָה־לִּי Nu 237, see §67o; (c) in the two plural forms קָֽדָשִׁים sanctuaries and שָֽׁרָשִׁים roots (also written קֳד׳ and שֳׁר׳). In all these cases the Jewish grammarians regard the Metheg accompanying the ־ָ as indicating a Qāmeṣ raḥabh (broad Qameṣ) and therefore read the ־ָ as ā; thus pā-olô, dā-rebān, pā-ŏlekhā, qā-dāšîm. But neither the origin of these forms, nor the analogous formations in Hebrew and in the cognate languages, nor the transcription of proper names in the LXX, allows us to regard this view as correct. It is just possible that Qameṣ is here used loosely for å̄, as the equivalent of ō, on the analogy of פֹּֽעֲלוֹ &c., §93q. As a matter of fact, however, we ought no doubt to divide and read pŏʿo-lô (for pŏʿ-lô), pŏʿŏ-lekhā, qŏdā-ším.—Quite as inconceivable is it for Metheg to be a sign of the lengthening into ā in בָּֽחֳרִי־אָֽף (Ex 118), although it is so in בָּֽאֳנִי bā-ʾonî (in the navy), since here the ā of the article appears under the ב.
- Of a different kind are the cases in which א has lost its consonantal sound by coalescing with a preceding a, a §23a–d.
- In Arabic this ă is always retained in an open syllable.
- At least according to the Masoretic orthography; cf. Wellhausen, Text der Bb. Sam., p. 18, Rem.
- Jerome (cf. Siegfried, ZAW. 1884, p. 77) in these cases often gives ă for ĭ.
- Cf. the remarks of I. Guidi, ‘La pronuncia del ṣērē,’ in the Verhandl. des Hamburger Orient.-Kongr. of 1902, Leiden, 1904, p. 208 ff., on Italian e for Latin i, as in fede = fĭdem, pece = pĭcem.
- Cf. Delitzsch, Physiologie u. Musik, Lpz. 1868, p. 15 f.
- Cf. above, b, end. On Jerome’s transliteration of o for ā, see ZAW. 1884, p. 75.
- These statements, in order to be fully understood, must be studied in connexion with the theory of syllables (§ 26) and Metheg (§16c–i).
- In the Babylonian punctuation (§8g, note) ā and ŏ are carefully distinguished. So also in many MSS with the ordinary punctuation and in Baer’s editions of the text since 1880, in which ־ְֳ is used for ŏ as well as for ŏ. Cf. Baer-Delitzsch, Liber Jobi, p. 43. But the identity of the two signs is certainly original, and the use of ־ֳ for ŏ is misleading.