Verbs.—As with nouns, so with verbs, modifications of mean ing were expressed by changes in the vocalization or the addition of prefixes. Thus from the common root q-t(t)-1 Arabic formed qatala, "killed"; qattala, "massacred", qdtala, "fought with"; 'aqtala, "exposed to death"; taqattala, "was busied with killing"; taqdtala, "fought together"; inqatala, "was killed"; istaqatala, "killed himself (through over-exertion)." Assyrian had i 2, Hebrew and Aramaic seven, Arabic II, and each one or more rarer, themes; Ethiopic had altogether 23 themes. Classical Arabic alone has a passive formed from each theme by substituting u-i-a for a-a-a, as in qutila, "was killed," and so on; Hebrew exhibits such a passive in the intensive and causative themes, such as quddash, "was sanctified," from qiddash, "sanctified." Elsewhere, how ever, only the passive participle (e.g., Heb. ydliidh, Aram. yelidh and Eth. welid, "born") has normally survived. Canaanite, like Ass.-Bab., used the "permansive" tense in both senses, though with different vowels (e.g., laqihu "they were taken," and laqahu, "they took") ; but this died out, and everywhere except in the eastern group the reflexive was more and more used as a passive ; e.g., Heb. nilqah beside luqqah, "was taken," from 14'4, "took" (cf. Ass.-Bab. liqi, "was taken," and "took") and Syr. ethmli, "was filled," from mld, "filled." The basic form was the third person singular unaugmented by prefix or affix, originally a nominal form denoting a timeless state, neither active nor passive ; for example, Ass.-Bab. nadin, "given," whether "has given" or "has been given." Ass.-Bab. used it generally but not exclusively as intransitive or passive, al though this led to ambiguity; for kasap nadin meant either "the money has been given" or "he gave the money." The Eastern group, assigning both uses to this form (called "penman sive" as expressing a more or less permanent state), removed the ambiguity by employing case-endings; the Western dispensed with them, since it marked the transitive by a and the intransitive by i under the second consonant. A third form with u (e.g., Ass.-Bab. marus, "was ill," or Heb. qaton, "was small") had always only a "stative" force. Occasionally the other groups confused the i- and a- forms (cf. Heb. male' and Eth. mal'a = Ass.-Bab. mali, "is full" or "has filled") but generally distinguished them (cf. Arab. mali'a, "was full," and mala'a, "filled") ; although usually differ ent roots were preferred (e.g., Heb. hakham, "was wise," and shdphat, "judged," (cf. Arab. hakuma, "was wise," and hakama, "judged").
The pronominal subject was affixed to the permansive tense, as being properly nominal: thus kabit attd, "honourable (art) thou," gave rise to kabtdtd, "thou-art-honourable"; this tense denoted a state, viz., what was complete in past time. A new tense, describ ing incomplete action in present or future time, was obtained by prefixing the pronominal element: thus ta- (from attd), "thou," and zakir coalesced into tazakar (with a for i under k, as the sense is active), "thou dost" or "wilt remember." Finally, a tense describing incomplete action in past time was devised by modifying the vowels of the present, giving rise to tazkur, "thou wast remembering," which came to serve for the pure preterite "thou didst remember." The cognate languages, through using the permansive as a proper preterite, required only an imperfect for incomplete action, past, present or future, for which they adopted a form resembling the Ass.-Bab. preterite (e.g., Heb. tizkor, "thou wast remembering," "dost" or "wilt remember," but zakhartd, "thou bast remembered"). Why a form corresponding with the Ass.-Bab. izakar, "he does" or "will remember," was not adopted, is not clear; it may have been to avoid confusion with the pret erite in certain derived themes. In some Aramaic dialects, espe cially Syriac and Mandaic, 1- or n- was prefixed to the third per son. In prohibitions, not the imperative but a tense resembling the preterite-imperfect (probably derived independently from the imperative) is used ; cf. Ass.-Bab. zukur and Heb. zekhor, "re member," but Ass.-Bab. a tazkur and Heb. 'al tizkar, "do not remember." Much skill was shown in circumventing the inadequacy of these tenses, notably by means of "wdw-consecutive" in Hebrew (q.v.) and the active participle in all the languages. But the need of com pound tenses was soon felt. Phoenician, Aramaic and Syriac formed with the verb "to be" a true present, a pluperfect, and other tenses; modern Arabic uses it as well as rdyih, "going," `ammell, "doing," and so on, for a similar purpose, and prefixes bi-, "in," to the imperfect to make a present (e.g., yiktub, "he
writes," and byiktub, "he is writing"). Mishnaic Hebrew formed a present by combining the participle with the pronominal sub ject (e.g., 'omerani, "I say," formed from 'Omer 'ani, "I [am] say ing"). The modem Syriac of Kurdistan has entirely discarded the old in favour of new tenses like this participial present and the participle with the verb "to be" and other formative elements.
All, especially Hebrew, being relatively deficient in compound nouns and adjectives, use numerous circumlocutions for them; e.g., Arab. qillat sabr, "littleness of patience—impatience"; all again, especially Arabic, frequently use apposition or a genitive case for an adjective; e.g., Arab. issitat addahab, "the watch, the gold," or scrat addahab, "the watch of gold." Arabic alone has a special form for the comparative and superlative of adjectives, of which Hebrew preserves traces ; all use a partitive preposition to express comparison ; e.g., Heb. "small from my brothers" = E. "smaller than my brothers." Most, but especially Hebrew and Arabic, lack adverbs, which they replace by the adverbial accusa tive and prepositional phrases. The syntax of all is very alike; but the Eastern languages when influenced by Sumerian tend to throw the verb to the end, the Western to the beginning of the sentence. All employ nominal clauses owing to difficulties about the verb "to be" and the absence (except in Assyrian) of a verb "to have" (e.g., Arab. "with me money" = E. "I have money"). Equally common is the circumstantial clause, viz., a clause added, often asyndetically, to describe the circumstances (e.g., Heb. "a tower and its head in Heaven"=E. "a tower whose top may reach unto Heaven"). Arabic is peculiar in putting the predicate in such a clause and after several verbs, notably the substantive, in the accusative case (e.g., Arab. "he went out, his father [nom.] sitting" [acc.] = E. "while his father was seated"). The verb, espe cially in Hebrew and Arabic, when at the beginning of the sen tence, often stands in the third person masculine singular (cf. Gk. schema Pindaricum). Co-ordinating and to a greater de gree sub-ordinating conjunctions, especially in Hebrew, were rare; Syriac, however, by borrowing largely from Greek (e.g., Syr. ger=Gk. 76,p, "for") obtained a more flowing construction of sentences. Otherwise )4L.s "speech strung together," chiefly with the conjunction "and," is characteristic of Semitic idiom (e.g., Arab. "he was forced and did it" =E. "forced to do it"). This conjunction frequently also introduces the apodosis (e.g., Heb. "if I find and I will spare"=E. "then I will spare"). So Ass.-Bab. used "and let" for "that" (e.g., Capp. "see and let thy instructions come" =E. "that thy instructions come"). Subordinate clauses were not generally distinguished; but Ass. Bab. marked verbs in them by the ending -u or sometimes -a and Arabic by -a. Ethiopic alone had a special form, using yezaker, "he remembers," as an imperfect and yezker, "let him remember," as a subjunctive. The corresponding form in Hebrew, though an imperfect, served also, if possible, in a shortened form, as a jussive (e.g., Heb. yizkor, "let him remember") and in Ass.-Bab. though a preterite, with /fi, "let," as a precative (e.g., Ass.-Bab. lizkur, "let him remember"). The infinitive in the accusative case, some times with an epithet, was often added to strengthen the verbal idea, especially in Arabic ; in Babylonian a primarily exclamatory infinitive in the nominative case served the same purpose (e.g., Ass.-Bab. tabalum tatbal, "a taking away! hast thou taken away?" =E. "hast thou indeed taken away?"). Hebrew uses construc tions apparently similar to both.
The Semitic languages, therefore, shared many common char acteristics ; but each, although all stand more closely together than any modern languages, developed peculiarities, which appear sporadically in the others. Thus Ass.-Bab. shared some words or idioms with Hebrew, others with Syriac ; Hebrew in some things resembled Phoenician or Moabite, in others Aramaic; Aramaic here touched Phoenician, there Arabic ; Arabic had much in com mon with Ethiopic, which in some points approached Accadian and Hebrew rather than any other language. See ASSYRIAN, HEBREW, and SYRIAC LANGUAGE; ETHIOPIC LITERATURE; AMHARIC; ARABIC and YIDDISH LANGUAGE. (G. R. D.)