Gesenius' Hebrew Grammar/23. The Feebleness of the Gutturals א and ה
a 1. The א, a light and scarcely audible guttural breathing, as a rule entirely loses its slight consonantal power whenever it stands without a vowel at the end of a syllable. It then remains (like the German h in roh, geh, nahte) merely as a sign of the preceding long vowel, e.g. מָצָא, מָלֵא, הוֹצִיא (but when a syllable is added with an introductory vowel, according to b below, we have, e.g. מְצָאַ֫נִי, הֽוֹצִיאַ֫נִי, since the א then stands at the beginning of the syllable, not מְצָאנִי, הוֹצִיאנִי), מְצֹא, כָּלוּא (cf., however, § 74 a), מָצָ֫אתָ (for māṣaʾtā), תִּמְצֶ֫אנָה. Similarly in cases like חֵטְא, וַיַּרְא, שָׁוְא, &c. (§ 19 l), and even in דֶּ֫שֶׁא, פֶּ֫לֶא (see above, § 22 e), the א only retains an orthographic significance.
b 2. On the other hand, א is in general retained as a strong consonant whenever it begins a syllable, e.g. אָמַר, מָֽאֲסוּ, or when it is protected by a Ḥaṭeph after a short syllable, e.g. לֶֽאֱכֹל, and finally, when it stands in a closed syllable with quiescent Šewâ after a preceding Seghôl or Pathaḥ, e.g. וַיֶּאְסֹר, נֶאְדָּר näʾdār, יַאְדִּ֫ימוּ yaʾdîmû. Even in such cases the consonantal power of א may be entirely lost, viz.
c (a) when it would stand with a long vowel in the middle of a word after Šewâ mobile. The long vowel is then occasionally thrown back into the place of the Šewâ, and the א is only retained orthographically, as an indication of the etymology, e.g. רָאשִׁים heads (for reʾāšîm), מָאתַ֫יִם two hundred (for meʾātháyim), שָֽׁאטְךָ Ez 256 for שְׁאָֽטְךָ; בּוֹדָאם Neh 68 for בּֽוֹדְאָם; מאוּם Jb 317, Dn 14 for מְאוּם; פֻּארָה for פְּאֻרָה Is 1033; חֹטִאים ḥôṭîm, 1 S 1433 for חֹֽטְאִים (cf. § 74 h, and § 75 oo); הָֽראוּבֵנִי Nu 3414, from רְאוּבֵן; so always חַטֹּאת or חַטֹּאות 1 K 1416, Mi 15, &c., for חַטְּאוֹת. Sometimes a still more violent suppression of the א occurs at the beginning of a syllable, which then causes a further change in the preceding syllable, e.g. מְלָאכָה work for מַלְאָכָה (as in the Babylonian punctuation), יִשְׁמָעֵאל for יִשְׁמַעְאֵל; שְׂמֹאל or שְׂמֹאול the left hand, ground form simʾâl.
d (b) When it originally closed a syllable. In these cases א is generally (by § 22 m) pronounced with a Ḥaṭeph, ־ֲ or ־ֱ. The preceding short vowel is, however, sometimes lengthened and retains the following א only orthographically, e.g. וַיָּ֫אצֶל Nu 1125 for וַיַּֽאֲצֵל (cf. Ju 941), and פָּארוּר Jo 26 for פַּֽאֲרוּר; לֵאמֹר for לֶֽאֱמֹר; לִֽאלֹהִים for לֶֽאֱלֹהִים; but the contraction does not take place in לֶֽאֱלִילֶ֫יהָ Is 1011. The short vowel is retained, although the consonantal power of א is entirely lost, in וַֽאדֹנָי, &c. (see § 102 m), וַיַּאת Is 4125, וָֽאַבֶּדְךָ Ez 2816 for וָֽאֲאַבֶּדְךָ; cf. Dt 2410, 1 K 1139, Is 1013.
e Instead of this א which has lost its consonantal value, one of the vowel letters ו and י is often written according to the nature of the sound, the former with ô and the latter with ô and î, e.g. רֵים buffalo for רְאֵם. At the end of the word ה also is written for א, יְמַלֵּה he fills for יְמַלֵּא Jb 821 (see below, l).
f 3. When א is only preserved orthographically or as an indication of the etymology (quiescent), it is sometimes entirely dropped (cf. § 19 k), e.g. יָצָ֫תִי Jb 121 for יָצָ֫אתִי; מָלֵ֫תִי Jb 3218 for מָלֵאתִי; מָצָ֫תִי Nu 1111; וַתֹּ֫הֶז 2 S 209; וַיְרַפּוּ Jer 811 for וַיְרַפְּאוּ; וַתַּזְּרֵ֫נִי 2 S 2240, but וַתְּאַזְּרֵנִי ψ 1840; תּוֹמִם Gn 2524 for תְּאוֹמִם; אֲחַטֶּנָּה 3139, for , אֲחַטְּאֶנָּה; שֵֽׁלָתֵךְ 1 S 117 for שְׁאֵֽלָ״; רֵמִים ψ 2222 for רְאֵמִים; גֵּוָה Jb 2229 for גְּאֵוָה; הַבֵּֽרֹתִי 1 Ch 1139 for הַבְּאֵֽר״, and so 2 S 2337; שֵׁרִית 1 Ch 1238 for שְׁאֵרִית; לַהְשׁוֹת 2 K 1925 Kethîbh for לְהַשְׁאוֹת (cf. Is 3726); חֵמָה Jb 296 for חֶמְאָה. In מַכֹּ֫לֶת 1 K 525 (for מַֽאֲכ״) the strengthening of the following consonant by Dageš compensates for the loss of the א; in מָסֹ֫רֶת Ez 2037, if for מַֽאֲס״ (but read מוּסָר, with Cornill), the preceding vowel is lengthened; cf. above, c. On אֹמַר for אֹאמַר, see § 68 g.
g Rem. 1. In Aramaic the א is much weaker and more liable to change than in Hebrew. In literary Arabic, on the other hand, it is almost always a firm consonant. According to Arabic orthography, א serves also to indicatea long a, whereas in Hebrew it very rarely occurs as a mere vowel letter after Qameṣ; as in קָאם Ho 1014 for קָם he rose up; רָאשׁ Pr 104, 1323 for רָשׁ poor; but in 2 S 111 the Kethîbh הַמַּלְאָכִים the messengers, is the true reading; cf. § 7 b.
h 2. In some cases at the beginning of a word, the א, instead of a compound Šeuâ, takes the corresponding full vowel, e.g. אֵזוֹר girdle for אֱזור; cf. § 84a q, and the analogous cases in § 52 n, § 63 p, § 76 d, § 93 r (אֽהָלְים).
i 3. An א is sometimes added at the end of the word to a final û, î, or ô, e.g. הָֽלְכוּא for הָֽלְכוּ Jos 1024 (before !א), אָבוּא Is 2812. These examples, however, are not so much instances of ‘Arabic orthography’, as early scribal errors, as in יִנָּשׂוּא Je 105 for יִנָּֽשְׂאוּ; and in נָשׂוּא ψ 13920 for נָֽשְׂאוּ. Cf. also יְהוּא Ec 113 (§ 75 s); נָקִיא for נָקִי pure; לוּא for לוּ if; אֵפוֹא for אֵפוֹ then (enclitic); רִבּוֹא for רִבּוֹ myriad, Neh 766.71. On הוּא and הִיא see § 32 k.
k 4. The ה is stronger and firmer than the א, and never loses its consonantal sound (i.e. quiesces) in the middle of a word except in the cases noted below, in which it is completely elided by syncope. On the other hand, at the end of a word it is always a mere vowel letter, unless expressly marked by Mappîq as a strong consonant (§ 14 a). Yet at times the consonantal sound of הּ at the end of a word is lost, and its place is taken by a simple ה or more correctly הֿ, with Rāphè as an indication of its non-consonantal character, e.g. לָהֿ to her for לָהּ, Zc 511, &c. (cf. § 103 g, and §§ 58 g, 91e); cf. also יָה for יָהּ (from יָהוּ) in proper names like יִרְמְיָה, &c.—Finally, in very many cases a complete elision of the consonantal ה takes place by syncope: (a) when its vowel is thrown back to the place of a preceding Šewâ mobile (see above, c, with א), e.g. לַבֹּ֫קֶר for לְהַבֹּ֫קֶר (the ה of the article being syncopated as it almost always is); כַּיּוֹם for כְּהַיּוֹם [but see § 35 n], בֲּשָּׁמַ֫יִם for בְּהַשָּׁמַ֫יִם; יֽוֹנָתָן for יְהֽוֹנָתָן; perhaps also בְּנִיהֶם for בְּנְהִיהֶם Ez 2732. (b) By contraction of the vowels preceding and following the ה, e.g. סוּסוֹ (also written סוּסֹה) from sûsahu (a+u=ô).—A violent suppression of ה together with its vowel occurs in בָּם (from בָּהֶם), &c.
l Rem. In connexion with ō and ē, a ה which only marks the vowel ending is occasionally changed into ו or י (רָאוֹ=רָאֹה, חַכֵּי=חַכֵּה Ho 69), and with any vowel into א in the later or Aramaic orthography, but especially with ā, e.g. שֵׁנָא sleep, ψ 1272 for שֵׁנָה; נָשֹׁא Jer 2339 for נָשֹׁהּ, &c. Thus it is evident that final ה as a vowel letter has only an orthographical importance.
- In Jer 2223, נֵחַנְתְּ is unquestionably a corruption of ננחת for נֶֽאֱנַ֫חַתְּ.
- Only apparent exceptions are such proper names as עֲשָׂהאֵל, פְּדָהצוּר, which are compounded of two words and hence are sometimes even divided. Cf. forms like חֲזָאֵל for חֲזָהאֵל. Another exception is יְפֵהפִיָּה, the reading of many MSS. for the artificially divided form יְפֵה־פִיְּה in the printed texts, Je 4620.