Gesenius' Hebrew Grammar/19. Changes of Consonants

§19. Changes of Consonants.

a The changes which take place among consonants, owing to the formation of words, inflexion, euphony, or to influences connected with the progress of the language, are commutation, assimilation, rejection, addition, transposition, softening.

1. Commutation[1] may take place between consonants which are either homorganic or homogeneous (cf. § 6 q), e.g. עָלַץ, עָלַס, עָלַז to exult, לָאָה, לָהָה, Aram. לְעָא to be weary, לָחַץ and נָחַץ to press, סָגַר and סָכַר to close, מָלַט and פָּלַט to escape. In process of time, and partly under the influence of Aramaic, the harder and rougher sounds especially were changed into the softer, e.g. צָחַק into שָׂחַק to laugh, גָּעַל into גָּאַל to reject, and the sibilants into the corresponding mutes: ז into ד, שׁ into ת, צ into ט. In many cases these mutes may be regarded as a return to an earlier stage of the pronunciation.

The interchange of consonants, however, belongs rather to the lexicographical treatment of stems[2] than to grammatical inflexion. To the latter belong the interchange (a) of ת and ט in Hithpaʿēl (§ 54 b); (b) of ו and י in verbs primae Yôd (§ 69), יָלַד for וָלַד, &c.

b 2. Assimilation usually takes place when one consonant which closes a syllable passes over into another beginning the next syllable, and forms with it a strengthened letter, as illustris for inlustris, affero for adfero, συλλαμβάνω for συνλαμβάνω. In Hebrew this occurs,

c (a) most frequently with נ‍, e.g. מִשָּׁם (for min-šām) from there, מִזֶּה (for min-zè) from this, יִתֵּן (for yintēn) he gives. נ‍ is not assimilated after the prefix לְ, e.g. לִנְגֹּף, nor as a rule before gutturals (except sometimes before ח), nor when it is the third consonant of the stem, e.g. שָׁכַ֫נְתָּ (cf. however נָתַ֫תָּ for nāthántā) except when another Nun follows, cf. § 44 o; nor in some isolated cases, as Dt 339, Is 291, 583, all in the principal pause; on הִנְדֹּף and תִּנְדֹּף ψ 683, see § 51 k, and § 66 f.

d (b) Less frequently and only in special cases with ל, ת, ד, e.g. יִקָּח (for yilqaḥ) he takes; מִדַּבֵּר for mithdabbēr; יִטַּמָּא for yithṭammā; תִּכּוֹנֵן for tithkônēn; תִּנַּשֵּׂא for תִּתְנַשֵּׂא; אַחַ֫ת for ʾaḥadt; but in 1 S 419 for לָלַת read probably לָלֶ֫דֶת.

e (c) In isolated cases with ה, ו, י, e.g. אָֽנָּא prithee! if from אָהּ נָא; ו and י mostly before sibilants in the verbal forms enumerated in § 71.

f In all these cases, instead of the assimilated letter, a Dageš forte appears in the following consonant. Dageš, however, is omitted when the strengthened consonant would stand at the end of a word, since the strengthening would then be less audible (§ 20 l), e.g. אַף nose (from ʾanp), תֵּת to give (from tint).

The cases are less frequent where a weak letter is lost in pronunciation,[3] and in place of it the preceding stronger sound is sharpened, i.e. takes Dageš, e.g. קְטָלַ֫תּוּ from קְטָלַ֫תְהוּ (§ 59 g). אֶסַּק for אֶסְלַק (§ 66 e) is an Aramaism.

g 3. Complete rejection takes place only in the case of weaker consonants, especially the sonants נ and ל, the gutturals א and ה, and the two half vowels ו and י. Such rejection takes place,

h (a) at the beginning of a word (aphaeresis), when these weak consonants (א, י, ל, נ‍) are not supported by a full vowel, but have only Šewâ, e.g. נַ֫חְנוּ we, also אֲנַ֫חְנוּ; דַּע for וְדַע; קַח for לְקַח; גַּשׁ for נְגַשׁ, הִי for נְהִי Ez 210.

i Aphaeresis of a weak consonant with a full vowel is supposed to occur in רַד Ju 1911 for יָרַד; in תַּ֫תָּה 2 S 2241 for נָתַ֫תָּה; in שׁוֹב for יָשׁוֹב Je 4210; on קָח Ez 175 for לָקַח, and on קָחָם Ho 113 for לְקָחָם, see § 66 g, end. In reality, however, all these forms are to be regarded merely as old textual errors.

k (b) In the middle of a word (syncope), when Šewâ precedes the weak consonant[4]; thus in the case of א (see further § 23 b–f, and § 68 bk), e.g. in מוּם for מְאוּם. As a rule in such cases, however, the א is orthographically retained, e.g. לִקְרַאת for לְקִרְאַת. Syncope occurs frequently in the case of ה, e.g. לַמֶּ֫לֶךְ for לְהַמֶּ֫לֶךְ (§ 23 k and § 35 n), יַקְטִיל for יְהַקְטִיל (§ 53 a).

Syncope of א with Šewâ occurs in such cases as בַּֽאדֹנָי for בַּֽאֲדֹנָי (cf. § 102 m); וַאעְשִׁר Zc 115.[5] On the cases in which א is wholly omitted after the article, see § 35 d.

Finally, the elision of ו and י in verbs ל״ה (§ 75 h) is an instance of syncope.—On the syncope of ה between two vowels, see § 23 k.

l (c) At the end of a word (apocope), e.g. גִּלֹה pr. name of a city (cf. גִּֽילֹנִי Gilonite); וַיַּרְא, where א though really rejected is orthographically retained, &c. On the apocope of ו and י in verbs ל״ה, see § 24 g, and § 75 a.

Bolder changes (especially by violent apocope), took place in earlier periods of the language, notably the weakening of the feminine ending ־ַת ăth to ־ָה ā, see § 44 a, and § 80 f.

m 4. To avoid harshness in pronunciation a helping sound, Aleph prosthetic[6] with its vowel, is prefixed to some words, e.g. אֶזְרוֹעַ and זְרוֹעַ arm (cf. χθές, ἐχθές; spiritus, French esprit).—A prosthetic ע occurs probably in עַקְרָב scorpion; cf. Arab. ʿuṣfûr bird (stem ṣafara).

n 5. Transposition[7] occurs only seldom in the grammar, e.g. הִשְׁתַּמֵּר for הִתְשַׁמֵּר (§ 54 b) for the sake of euphony; it is more frequent in the lexicon (כֶּ֫בֶשׂ and כֶּ֫שֶׂב lamb, שִׂמְלָה and שַׂלְמָה garment), but is mostly confined to sibilants and sonants.

o 6. Softening occurs e.g. in כּוֹכָב star, from kaukabh=kawkabh for kabhkabh (cf. Syriac raurab=rabrab); טֽוֹטָפוֹת phylacteries for ṭaphṭāphôth; according to the common opinion, also in אִישׁ man from ʾinš, cf. however § 96.

  1. Cf. Barth, Etymologische Forschungen, Lpz. 1893, p. 15 ff. (‘Lautverschiebungen’).
  2. See in the Lexicon, the preliminary remarks on the several consonants.
  3. Such a suppression of a letter is sometimes inaccurately called ‘backward assimilation’.
  4. Syncope of a strong consonant (ע) occurs in בִּי prithee! if this stands for בְּעִי (see Lexicon), also in ונשׁקה Am 88, Kethîbh for וְנִשְׁקְעָה (cf. וְשָֽׁקְעָה 95), and in בָּלָה Jos 193 for בָּֽעֲלָה (as in 1529). Probably, however, ונשׁקה and בלה are only clerical errors, as is undoubtedly כָאֹר Am 88 for כַיְאֹר (95).
  5. Frensdorff, Ochla Wʾochla, p. 97 f., gives a list of forty-eight words with quiescent א.
  6. This awkward term is at any rate as suitable as the name Alef protheticum proposed by Nestle, Marginalien u. Materialien, Tübingen, 1893, p. 67 ff.
  7. Cf. Barth, Etymologische Studien, Lpz. 1893, p. 1 ff.; Königsberger, in Zeitschrift f. wissenschaftliche Theologie, 1894, p. 451 ff.