596 Derivation from the roots and inflexion proceed partly by the reduplication of root letters and the addition of cer tain preformatives and affonnatives (more rarely by the insertion af formative consonants in the body of the root), partly by nudiflcations of the vowels with which the radi cals are pronounced. Almost every root expresses in its origin something that can be grasped by the senses, and the mechanism by which words are formed from the root is adapted te present sensible notions in a variety of nuances and in all possible embodiments and connexions, so that there are regular forms to express in a single word the in tensity, the repetition, the production of the root idea the place, the instrument, the time of its occurrence, and so forth. Thus the expression of intellectual ideas is neces sarily metaphorical, almost every word being capable of a material sense or at least conveying the distinct suggestion of some sensible notion. For example, the names of passions depict their physiological expression ; " to confer honour" means also "to make heavy," and so on. The same concrete character, the same inadequacy to convey purely abstract thoughts without a substratum appealing to the senses, appears in the grammatical structure of the Semitic tongues, for example, in the absence of the neuter gender, in the extreme paucity of particles, in the scanty provision for the subordination of propositions, which de prives the Semitic style of all involved periods and reduces it to a succession of short sentences linked by the simple copula and. The fundamental element of these languages is the noun, and in the fundamental type of sentence the predicate is a noun set down without any copula and there fore without distinction of past, present, or future time. The finite verb is developed from nominal forms (participial or infinitive), and is equally without distinction of time. Instead of tenses we find two forms, the perfect and the imperfect, which are used according as the speaker con templates the verbal action as a thing complete or as con ditional, imperfect, or in process. It lies in the nature of this distinction that the imperfect alone has moods. In their later stages the languages seek to supply the lack of tenses by circumlocutions with a substantive verb and par ticiples. Other notable features common to the Semitic tongues are the use of appended suffixes to denote the possessive pronouns with a substantive, or the accusative of a personal pronoun with a verb, and the expression of the genitive relation by what is called construction or an nexation, the governing noun being placed immediately before the genitive, and if possible slightly shortened in pronunciation so that the two words may run together as one idea. A characteristic of the later stages of the lan guages is the resolution of this relation into a prepositional clause. These and other peculiarities are sufficient to establish the original unity of the group, and entitle us to postulate an original language from which all the Semitic dialects have sprung. Of the relation of this language to other linguistic stems, especially to the Indo-Germanic on the east and the North-African languages on the west, we cannot yet speak with certainty ; but it appears that the present system of triliteral roots has grown out of an earlier biliteral system which, so far as it can be recon structed, must form the basis of scientific inquiry into the ultimate affinities of the Semitic group. 1 Before the rise of comparative philology it was a familiar opinion that Hebrew is the original speech of mankind. Taken from the Jews, and as already expressed in the Palestinian Targum on Gen. xi. 1, this opinion drew its 1 Renan, Ilistoire des Langues Semitiques, sketches the history of research in this direction. Noteworthy are the remarks of Lagarde, fiymmicta, p. 121. On survivals from the biliteral stage, see Noldeke, Mandaische Gram., p. 96. main support from etymologies and other data in the earlier chapters of Genesis, which, however, were as plausibly turned by Syriac writers in favour of their own tongue. 2 Till quite recently many excellent scholars (including Ewald) have claimed for Hebrew the greatest relative an tiquity among Semitic tongues. But though Hebrew has by far the oldest literature, this does not prove that its structure comes nearest to the original Semitic language. And it is now generally recognized that in grammatical structure the Arabic, shut up within its native deserts till the epoch of Islam, preserved much more of the original Semitic forms than either Hebrew or Aramaic. In its richer vocalization, in the possession of distinct case-end ings, in the use for feminine nouns of the afformative f, which in the northern dialect has passed through h (originally audible as in Egyptian Arabic) into a mere vowel, in the more extensive range of passive and modal forms, and in other refinements of inflexion, Arabic repre sents no later development, but the original wealth and primitive subtlety of Semitic speech, as appears not only from fragmentary survivals in the other dialects but from an examination of the process of decay which has brought the spoken Arabic of the present day into a grammatical condition closely parallel to the Old Testament Hebrew. But while Arabic is in many respects the elder brother, it is not the parent of Hebrew or Aramaic. Each member of the group had an independent development from a stage prior to any existing language, though it would seem that Hebrew did not branch off from Aramaic so soon as from Arabic, while in its later stages it came under direct Aramaic influence. Among the points in which Hebrew di tiers both from Arabic and Aramaic may be mentioned the consonantal relations already spoken of, the system of tone long vowels, the use of Waw consecutive, the H of the causative and reflexive stems and of the article, and the compensation for omitting the reduplication of gutturals and r by lengthening a preceding vowel. 3 Again, Hebrew agrees with Arabic against Aramaic in possessing a prefixed article, in the use of the Nifal (Arab. vn. ), and in other minor points. But in more notable features Hebrew and Aramaic agree against Arabic, as in the absence of broken plurals, the place of the accent, the aspir ation of certain letters whenever they are preceded by a vowid sound, and the substitution of y for w as the first radical of roots. To give further details would carry us too far into comparative grammar. Speaking generally, it may be said that Hebrew is less copious in vocabulary than Arabic (in which tongue, however, the abundance of synonyms is largely due to an artificial combination of several dialects), and less rich in subtle distinctions of gram matical form and refinements of syntax. On the other hand, Hebrew is much superior to Aramaic in flexibility of structure, in fulness of vowels, and in all the qualities which adapt a language for poetical expression. Geographical /Sphere and History of Hebrew as a Spoken Laiifluaye.Tlie Hebrew spoken by the Israelites in Canaan was separated only by very minor differences (like those of our provincial dialects) from the speech of neigh bouring tribes. We know this for the Moabite language from the stone of Mesha ; and the indications furnished by proper names, as well as the acknowledged affinity of Israel with these tribes, make the same thing probable for Ammon and Edom. More remarkable is the fact that the Phoenicians and Canaanites, with whom the Israelites acknowledged no brotherhood, spoke a language which, at 8 Theodoret (Q)icest. in Gen.}, xi., Barhebrocus, and others cited by Assemani, Bib. Or., iii. 314. The same opinion appears among the Babylonian Jews (Rab in Synh., 38b). Conversely, Jacob of Sarug concedes the priority of Hebrew (see Z. D. M. G., xxv.,. p. 520). The Arabs, whose language is in many points older than either, yield priority to Hebrew (Abulfeda, //. A., p. IS), or to Syriac (Tabari, i., 220 ; Abu Isa in Abulfeda, p. 148), the language of the race to which they owed their first knowledge of letters. 3 ^The verbal prefixes ha, hith, appear in Palestinian Aramaic, and the^Aramaic dialects also exhibit some cases of a lengthened vowel before r (Noldeke, Mand. Gram., p. 17). On the other hand, the
transcription of the LXX. shows that Hebrew once could double r.