appendix A Concise Historical Morphology of Biblical Hebrew We have considered the most important sound changes from Proto-Northwest-Semitic to Biblical Hebrew. This appendix reviews the consequences of these developments for the historical morphology of the inflected word classes of Biblical Hebrew: pronouns, nouns and adjectives, numerals, and verbs. Frequent reference is made to the chapters discussing individual sound changes. 1 Pronouns The personal pronouns, both independent and suffixed, are discussed in detail in Chap- ter 8. :derives from the old genitive *ḏī ֶזה The masculine near demonstrative pronoun word-final stressed *-ī > *-ē and word-final *-ē > *-ɛ̄. The associated nominative *ḏū and the accusative *ḏā developed to the זוּ regularly yields the rare relative pronoun ,is difficult to reconstruct ֵ֫אֶלּה due to the Canaanite Shift.The plural זוֹ/זֹה demonstrative with different Semitic languages showing irreconcilable forms.The Hebrew form seems to go back to *ʔillay: the unstressed diphthong contracted to *-ē and then participated in the shift to *-ɛ̄ in word-final position. , ֶזה < does not show the same shift as *ḏī ִמי As the animate interrogative pronoun it should not be reconstructed with word-final *-ī, but as *mīya. The inanimate inter- with gemination of the following consonant) goes ַמה־ normally) ָמה rogative pronoun back to *mah. © koninklijke brill nv, leiden, 2020 | doi:10.1163/9789004390263_011 Benjamin D. Suchard - 9789004390263 Downloaded from Brill.com09/30/2021 07:59:23AM via free access 232 appendix 2 Nouns and Adjectives1 2.1 Roots with One or Two Radicals also used before , ִפּי mouth’ has only one radical. The construct state‘ ֶפּה ,Synchronically suffixes, derives from *pī, originally the construct state genitive.2 This was either gen- eralized to the absolute state, where stressed word-final *-ī shifted to *-ē and fur- ther to *-ɛ̄, or the absolute state regularly developed from *pīm > *pím > *pém > *pḗ. man’ (only attested in the plural) should‘ * ֵמת bottom’, and‘ ֵשׁת ,’name‘ ֵשׁם ,’son‘ ֵבּן be reconstructed as biradical *qṭum nouns without a vowel between the first and sec- ond radical (Testen 1985, Blau 2006).3 The loss of mimation and the case vowel in the singular was probably analogous to their deletion in all other nominals. In the unsuf- fixed singular, an epenthetic vowel *e was inserted to break up the consonant cluster. his‘ ְשׁמוֹ ,’my son‘ ְבִּני The lack of an original vowel can still be seen before suffixes, as in -forms the irreg ֵבּן .men’, all without pretonic lengthening‘ ְמִתים name’, and in the plural banīma, a form which is at least Proto-West-Semitic (cf. its reflexes* > ָבִּנים ular plural has extended the singular stem to the ֵשׁם in Arabic and Modern South Arabian), and . ֵשׁמוֹת ,plural Nouns with a short vowel between their first and second radical undergo tonic and ֶשׂה < hands’. *śayum‘ ָי ַ֫דִים < hand’, *yadayma‘ ָיד < pretonic lengthening, e.g. *yadum ‘sheep’ shows triphthong contraction and the subsequent development of word-final ,’father‘ ָאב ē > *-ɛ̄. No unambiguous *quṭum nouns are attested. The irregular nouns-* father-in-law’ extend their stem with -i in the construct state and‘ ָחם brother’, and‘ ָאח ’brothers‘ ַאִחים before suffixes, originally a lengthened genitive case vowel. The plural goes back to *ʔaḫīma > *ʔaḫḫīma with pretonic gemination (Bergsträsser 1918, 139–149); his brothers’.4‘ ֶאָחיו < * ַאָחיו before å in the suffix, the a regularly assimilates to ɛ, as in 1 In this section, I will make frequent reference to the recent overview of Hebrew nominal morphology by Huehnergard (2015), where the reader will also find many references to the extensive literature on this subject. I have generally not indicated the many points on which we are in agreement. 2 Huehnergard (2015, 31) reconstructs the absolute state stem as *piy-, while Militarev and Kogan (2005, 195–197) reconstruct *pay- or *pVw-. According to the rules for triphthong con- although *piy- is , ֶפּה traction given in Chapter 5, all of these reconstructions would also yield incompatible with the Classical Arabic accusative fā (construct), not **fiya. 3 Many other scholars reconstruct a short *i or *u in these nouns, e.g. Huehnergard (2015, 30– 31). ַאַחד one (f.) (pause)’ but no change in‘ ֶא ָ֑חת < * ַא ָ֑חת ,’(one (m.) (absolute‘ ֶאָחד < * ַאָחד .Cf 4 lying‘ ֶכָּחִשׁים < * ַכָּחִשׁים ;’charcoal‘ ֶפָּחם < * ַפָּחם ;’(one (f.) (context‘ ַאַחת ,’(one (m.) (construct‘ ;’(the sword (pause‘ ֶה ָ֑חֶרב < * ַה ָ֑חֶרב ;’(he will comfort himself (pause‘ ִיְתֶנ ָ֑חם < * ִיְתַנ ָ֑חם ;’(.m.pl) etc. No sequences of aHå (with H representing any guttural) are retained. Benjamin D. Suchard - 9789004390263 Downloaded from Brill.com09/30/2021 07:59:23AM via free access a concise historical morphology of biblical hebrew 233 If the stress is farther removed from the first syllable, the original single *ḫ was pre- ֲאֵחיֶכם < brothers (construct)’ (Num 27:10), *ʔaḫaykimma‘ ֲאֵחי < served, as in *ʔaḫay ‘your (m.pl.) brothers’ (Hos 2:3).5 Nouns with *ī and *ū retain their original vowel, while those with *ā change it to *ō > o due to the Canaanite Shift. ֶ֫דֶּלת .Some feminine biradical nouns have incorporated a third radical in the plural ’maidservant‘ ָאָמה ; ְדָּלתוֹת door’ was reanalyzed as a segolate, giving rise to the plural‘ ָחמוֹת sister’ and‘ ָאחוֹת The feminine nouns . ֲאָמהוֹת ,has a third radical *h in the plural ‘mother-in-law’ were originally formed by adding the feminine suffix *-at- to their asso- ciated masculines; due to the pre-Proto-Semitic loss of their third radical, presumably *w, *-at- was lengthened to *-āt-, which became *-ōt- due to the Canaanite Shift. This parallels the lengthening of the case vowels still seen in the masculine construct states .(father-in-law’ (only attested before a suffix‘ * ֲחִמי brother’ and‘ ֲאִחי 2.2 *qVṭl(at)um (segolates) After the loss of the case endings, *qVṭlum nouns contained a word-final consonant cluster in the singular absolute state. In nouns from strong roots, this was resolved at a relatively late point in the history of Hebrew by the insertion of an epenthetic vowel, usually *ɛ, but *i after *y and usually *a after gutturals.6 If the third radical was -path’ (Bauer and Lean‘ ֹ֫אַרח < h, ḥ, or ʕ, *ɛ regularly assimilated to *a, as in *ʔórɛḥ ֶ֫חֶדר room’ (absolute‘ ֲחַדר der 1922, 569). In a few words, a separate construct state like is attested. With Steiner (1976), we may attribute this to early insertion of ( ֵ֫חֶדר and an epenthetic vowel before resonants; the further development of these forms, pre- sumably to be reconstructed like *ḥadr > *ḥader, is then like that of construct states of *qaṭilum nouns, discussed below. The plural forms of these segolates generally developed in the same way: in the abso- -with reduction of the first vowel and pretonic length , ְקָטִלים < lute state, *qVṭalīma7 ening of the second one. *quṭlum nouns seem to have preserved short *o in the first when secondarily accented, this vowel is ;( ֫גֶֹּרן .threshing floors’ (sg‘ ֳגָּרנוֹת syllable, as in In the construct state and before heavy .( ֫קֶֹדשׁ .holies’ (sg‘ ָֽקָדִשׁים lengthened to ɔ, as in 5 It thus appears that the gemination in the other forms of the plural is a (pre-)Hebrew-internal development, making the connection with Akkadian /aḫḫū/ ‘brothers’ more tenuous (pace Huehnergard 2015, 35). ֫בֶֹּהן bride-price’ (and also‘ ֹ֫מַהר tent’ and‘ ֹ֫אֶהל Based on cognate evidence, it seems likely that 6 ‘thumb/big toe’?) were originally *qaṭlum nouns (Huehnergard 2015, 40, n. 37). Similarly to head’, they would then have lost their *h with compensatory‘ רֹאשׁ < II-ʔ forms like *raʔsum lengthening of *a and a subsequent change to *ō under the Canaanite Shift, e.g. *ʔahl- > *ʔāl- > *ʔōl-. The consonantal h would then have been reintroduced based on the spelling, as may also have happened in some II-ʔ nouns (see below). 7 With the insertion of *a in the plural stem, possibly an Afroasiatic feature (cf. Greenberg 1955). Benjamin D. Suchard - 9789004390263 Downloaded from Brill.com09/30/2021 07:59:23AM via free access 234 appendix suffixes, *qVṭalay yielded *qəṭəlē. The first vowel developed to i or a, depending on the surrounding consonants and was frequently analogically influenced by the singular (see Chapter 7); *quṭlum again preserved ɔ. The second šwå was deleted, as it occurred in an open syllable. In the singular, we find different developments of the vowel in the first syllable. in context, with assimilation of *a to the epenthetic ֶ֫קֶטל qaṭlum regularly developed to* seed’ shows that‘ ֶ֫זַרע < ɛ. The development of words like *ḏarʕum > *zárɛʕ > *zɛ́rɛʕ* this assimilation took place before that of the second *ɛ to *a before gutturals. In major Before .ָ֑קֶטל pause, *a was lengthened to *ā, which did not assimilate, resulting in pausal suffixes, *a was largely preserved, but sometimes analogically replaced by *i due to influence from the *qiṭlum paradigm. *qiṭlum regularly merged with *qaṭlum in the context form of the absolute state due to Philippi and Blau’s Laws (see Chapter 6). Thus, the regular Biblical Hebrew -righteousness’. As Blau’s Law post‘ ֶ֫צֶדק < as in *tṣidqum , ֶ֫קֶטל outcome of *qiṭlum is Before suffixes, *i .ֶ֑קֶטל dated major pausal lengthening, the regular pausal form is also the , ֶ֫קֶטל < was regularly preserved. Since *qaṭlum and *qiṭlum partly merged in *qaṭl two paradigms exerted a large amount of analogical influence on each other, resulting his grave’ for regular‘ ִקְברוֹ milkahu,8* > ** ִמְלכּוֹ his king’ for regular‘ ַמְלכּוֹ in forms like -riglum. *e < *i was also fre* > **ֶ֑רֶגל foot’ for regular‘ ָ֑רֶגל qabrahu, and pausal* > ** ַקְברוֹ quently restored in the absolute state of *qiṭlum nouns, based on analogies like *kaspṓ ‘his silver’ : *kásp ‘silver’ = *ʕeglṓ ‘his calf’ : *ʕégl ‘calf’. These analogically restored *qeṭl . ֵ֫קֶטל nouns then developed into 2.2.1 Weak Roots In an early change, syllable-final *ʔ was lost, with compensatory lengthening of the preceding vowel.