PART SECOND.
MORPHOLOGY.
Strong and weak roots. § 57. The large majority of all Semitic words, as is well known, are derived from roots which for the most part have three, but occasionally even four or more 'Radicals'. If the three radicals are firm consonants, the roots are then called Strong: but if one of the radicals is ܘ or ܝ (frequently appearing as a vowel), or if the due weight of the word is attained by the doubling of one of two firm radicals, then the roots are called Weak. On practical grounds we retain this method of treating roots, without insisting farther on the point that even with strong roots a radical is often demonstrably of quite recent origin, while on the other hand there is much variety in the origin of weak forms of the root, and while in many cases at least, the assumption of an original Waw or Yod as a radical, or that of a third radical with the same sound as the second, is a pure fiction. Thus we speak of roots primae ܘ or ܝ (פו׳, פי׳) [Pe Waw, Pe Yod] meaning those whose first radical is taken as W or Y; so of roots mediae ܘ or ܝ (עו׳, עי׳) [Ayin Waw, Ayin Yod], and tertiae ܝ and mediae geminatae (עע׳) [Lamed Yod, and Ayin doubled]. In addition we have frequently to deal specially with words of which ܐ is a radical; for this sound (cf. § 33 sqq.) undergoes many modifications. In like manner we have to treat of words which have n as the first letter of the root. The forms too, which have a guttural or an r as second or third radical, are, by reason of certain properties, brought occasionally into special notice.
Variation of weak roots. § 58. Weak roots vary a good deal in their weak letters. Thus חמם, חום, חמי, יהם (to which is added another secondary form חמת) are essentially modifications only of the same fundamental root, which means "hot". In particular, roots עו׳ and עע׳ are very closely related. Thus also in Syriac they very readily change into one another: the substantive belonging to פדד "to err" (Perf. ܦܱܕ, Impf. ܢܷܦܷ݁ܕ) is ܦܱܘܕ݁ܐ, as if from פוד; and along with the frequently occurring חנן "to pity" חון is found (Perf. ܚܳܢ, Impf. ܢܚܽܘܢ), and with כפף "to bend", כוף, &c.
Roots med. gem. § 59. Forms med. gem. in Syriac attain like weight with that of the strong forms, by doubling not the second radical, but the first, when it is possible, i. e. when a prefix ending in a vowel precedes it. Thus from גזז "to shear" ܐܱܓܷ݁ܙ aggez (answering to ܐܱܘܜܶܠ); ܐܷܓ݁ܘܿܙ eggoz (= ܐܷܩܜܘܿܠ); ܢܷܓ݁ܙܽܘܢ negzūn, properly negge̊zūn (= ܢܷܩܜܠܽܘܢ); ܬܱܚ̈ܒ݂ܳܢ "you (fem. pl.) love" (= ܬܱܩ̈ܜܠܴܢ, from חבב); ܡܱܚܡܳܐ "boiler" (from חמם "to warm"); ܡܱܥܠܴܐ, ܡܱܥܱܠܬ݁ܐ, ܡܱ̈ܥܴܠܷܐ "entrance", &c.
Yet in some nouns we find the general Semitic method,—i. e. the method of either directly or virtually doubling the third radical, even with the prefixes mentioned: thus ܡܚܰܜܳܐ "needle" (not ܡܱܚܜܳܐ); ܡܓ݂ܶܢܳܐ or ܡܓ݂ܰܢܳܐ (East-Syrian) "shield"; ܡܜܰܠܬ݂ܐ "a booth" (me̊ṭalthā, properly me̊ṭalle̊thā), pl. ܡ̈ܜܰܠܷܐ (me̊ṭallē); ܡܨܰܠܬ݂ܐ "sieve"; ܡܥܱܪܬ݂ܳܐ "a cave"; and ܡܬ݂ܘܿܡ, ܡ̈ܬ݂ܘܿܡܱܘܗ̄ܝ, &c., mostly used adverbially, "completion" (תמם), "continually".
Two l's stand beside each other like two different consonants[1] in ܡܱܡܠܠܐ "speech"; ܡܱܜܠܠܷܐ "cover, shelter" (§ 46); and the quadriliteral form ܩܽܘܒܠܠܴܐ "face". In these formations, however, the l is again dropped in the usual pronunciation (§ 29), so that in point of fact the regular form makes its appearance. Add the peculiar form ܐܱܝܠܷܠܐ, ܐܱܝܠܠܱܬ &c. "to lament"[2]. The following appear to be later formations: ܬܱܗܠܱܠܬܴܐ "mockery", from ܐܱܗܶܠ (הלל); and from ܐܷܬܚܰܢܱܢ, ܬܱܚܢܱܢܬܴ݁ܐ "a prayer". Thus, farther, regularly in the Ethpeel ܐܷܬܓܙܶܙ "was shorn" (as compared with ܓܰܙ "shore").
In Syriac too the second and third radicals, when identical, are always kept in separate existence, if a long vowel comes between them, in the course of the formation, e. g. ܚܢܺܝܢܳܐ "pardoned"; ܚܢܳܢܳܐ "favour", &c., as well as when the first of the two is itself doubled, e. g. ܐܷܬܚܰܢܱܢ ethḥannan "begged for pardon".
Quadriliteral roots. § 60. With roots of four radicals we also rank such as are demonstrably formed originally from roots of three radicals with well-known suffixes or prefixes, but which are treated in the language quite like quadriliteral forms, e. g. ܫܱܥܒܷܕ "to enslave", properly a causative form from ܥܒܱܕ; ܢܱܟ݂ܪܻܝ "to estrange", "to alienate", from ܢܽܘܟ݂ܪ݂ܝ "strange", from נכר, &c
Nouns and verbs. § 61. Nouns, properly so called (Substantives and Adjectives), and verbs, have in all respects such a form that they are subject to the scheme of derivation from roots composed of three or more radicals, although sufficient traces survive to show that this condition was not, throughout and everywhere, the original one. The only marked divergences in formation, however, are found on the one hand with the Pronouns (which originate partly in the welding together of very short fragments of words), and on the other hand with many old Particles. To these two classes, the Pronouns and Particles,—we must therefore assign a separate place, although both in conception and usage they belong to the Noun. The same treatment must be extended to the Numerals, which, to be sure, stand in form much nearer to the usual tri-radical formations.
Interjections. § 62. Overagainst all true words, or words that express some conception, stand the expressions of feeling—or the Interjections, which originally are not true words at all, but gradually enter,—at least in part,—into purely grammatical associations, and even serve to form notional words. Thus ܘܳܝ "woe!" is a mere exclamation of pain, and ܦܘܿܝ "fye!" one of detestation; but ܘܳܝ ܠܓܰܒܪܴܐ "woe to the man!" or ܦܘܿܝ ܡܶܢ ܓܰܒܪܴܐ "fye upon the man!" is already a grammatical association of words, and ܘܳܝܳܐ "the woe" is a regular noun.[3]
Such Interjections are ܐܘّ (§ 9), ܐܘܿܢ "O!" ܝܳܝ, ܝܳܐ "O!"; ܐܘܿܗ "Ah!"; ܐܷܗܶܐ "Ho! Ho!" (in mockery), &c. Also the demonstrative form ܗܳܐ "Here!" "Lo!", which is greatly employed in the formation of Pronouns and Adverbs, is to be regarded as originally an interjection.