Gesenius' Hebrew Grammar/30. Stems and Roots; Biliteral, Triliteral, and Quadriliteral
ETYMOLOGY, OR THE PARTS OF SPEECH
30a 1. Stems in Hebrew, as in the other Semitic languages, have this peculiarity, that by far the majority of them consist of three consonants. On these the meaning essentially depends, while the various modifications of the idea are expressed rather by changes in the vowels, e.g. עמק (עָמֵק or עָמֹק; the 3rd pers. sing. perf. does not occur) it was deep, עָמֹ֫ק deep, עֹ֫מֶק depth, עֵ֫מֶק, a valley, plain. Such a stem may be either a verb or a noun, and the language commonly exhihits both together, e.g. זָרַע he has sown, זֶ֫רַע seed; חָכַם he was wise, חָכָם a wise man. For practical purposes, however, it has long been the custom to regard as the stem the 3rd pers. sing. Perf. Qal (see § 43), since it is one of the simplest forms of the verb, without any formative additions. Not only are the other forms of the verb referred to this stem, but also the noun-forms, and the large number of particles derived from nouns; e.g. קָדַשׁ he was holy, קֹ֫דֶשׁ holiness, קָדוֹשׁ holy.
30b Sometimes the language, as we have it, exhibits only the verbal stem without any corresponding noun-form, e.g. סָקַל to stone, נָהַק to bray; and on the other hand, the noun sometimes exists without the corresponding verb, e.g. אֶ֫בֶן stone, נֶ֫גֶב south. Since, however, the nominal or verbal stems, which are not now found in Hebrew, generally occur in one or more of the other Semitic dialects, it may be assumed, as a rule, that Hebrew, when a living language, also possessed them. Thus, in Arabic, the verbal stem ʾăbınă (to become compact, hard) corresponds to אֶ֫בֶן, and the Aramaic verb negab (to be dry) to נֶ֫גֶב.
30c Rem. 1. The Jewish grammarians call the stem (i.e. the 3rd pers. sing. Perf. Qal) שֹׁ֫רֶשׁ root. Hence it became customary among Christian grammarians to call the stem radix, and its three consonants litterae radicales, in contradistinction to the litterae serviles or formative letters. On the correct use of the term root, see g. 30d 2. Others regard the three stem-consonants as a root, in the sense that, considered as vowelless and unpronounceable, it represents the common foundation of the verbal and nominal stems developed from it, just as in the vegetable world, from which the figure is borrowed, stems grow from the hidden root, e.g.
|Root:||מלך,||the indeterminate idea of ruling.|
|Verb-stem, מָלַךְ he has reigned.||Noun-stem, מֶ֫לֶךְ king.|
For the historical investigation of the language, however, this hypothesis of unpronounceable roots, with indeterminate meaning, is fruitless. Moreover, the term root, as it is generally understood by philologists, cannot be applied to the Semitic triliteral stem (see f).
30e 3. The 3rd sing. Perf. Qal, which, according to the above, is usually regarded, both lexicographically and grammatically, as the ground-form, is generally in Hebrew a dissyllable, e.g. קָטַל. The monosyllabic forms have only arisen by contraction (according to the traditional explanation) from stems which had a weak letter (ו or י) for their middle consonant, e.g. קָם from qăwăm; or from stems whose second and third consonants are identical, e.g. צַר and צָרַר (but see below, §§67, 72). The dissyllabic forms have themselves no doubt arisen, through a loss of the final vowel, from trisyllables, e.g. קָטַל from qătălă, as it is in literary Arabic.
30f 2. The law of the triliteral stem is so strictly observed in the formation of verbs and nouns in Hebrew (and in the Semitic languages generally), that the language has sometimes adopted artificial methods to preserve at least an appearance of triliteralism in monosyllabic stems, e.g. שֶׁ֫בֶת for the inf. constr. of verbs פ״ו; cf. §69b. Conversely such nouns, as אָב father, אֵם mother, אָח brother, which were formerly all regarded as original monosyllabic forms (nomina primitiva), may, in some cases at least, have arisen from mutilation of a triliteral stem.
30g On the other hand, a large number of triliteral stems really point to a biliteral base, which may be properly called a root (radix primaria, bilitteralis), since it forms the starting-point for several triliteral modifications of the same fundamental idea. Though in themselves unpronounceable, these roots are usually pronounced with ă between the two consonants, and are represented in writing by the sign √, e.g. √כר as the root of כָּרַר, כָּרָה, כּוּר, אָכַר. The reduction of a stem to the underlying root may generally be accomplished with certainty when the stem exhibits one weak consonant with two strong ones, or when the second and third consonants are identical. Thus e.g. the stems דָּכַךְ, דּוּךְ, דָּכָא, דָּכָה may all be traced to the idea of striking, breaking, and the root common to them all is evidently the two strong consonants דך (dakh). Very frequently, however, the development of the root into a stem is effected by the addition of a strong consonant, especially, it seems, a sibilant, liquid or guttural. Finally, further modifications of the same root are produced when either a consonant of the root, or the letter which has been added, changes by phonetic laws into a kindred letter (see the examples below). Usually such a change of sound is accompanied by a modification of meaning.
30h Examples: from the root קץ (no doubt onomatopoetic, i.e. imitating the sound), which represents the fundamental idea of carving off, cutting in pieces, are derived directly: קצץ and קצה to cut, to cut off; the latter also metaph. to decide, to judge (whence קָצִין, Arab. qâḍi, a judge); also קָצַב to cut off, to shear, קָצַף to tear, to break, קָצַע to cut into, קָצַר to cut off, to reap. With a dental instead of the sibilant, קט, קד, whence קָטַב to cut in pieces, to destroy, קָטַל to cut down, to kill, קָטַף to tear off, to pluck off. With the initial letter softened, the root becomes כס, whence כָּסַח to cut off, and כָּסַם to shave; cf. also נכס Syr. to slay (sacrifice), to kill. With the greatest softening to גז and גד; גָּזַז to cut off, to shear; גָּזָה to hew stone; גּוּז, גָּזַם, גָּזַע, גָּזַל, גָּזַר to cut off, to tear off, eat up; similarly גָּדַד to cut into, גָּדַע to cut off; cf. also גָּדָה, גָּדַף, גָּדַר. Allied to this root also is the series of stems which instead of a palatal begin with a guttural (ח), e.g. חָרַד to split, cut; cf. also חדל, חדק, חדר, חדשׁ, and further חוּס, חוּץ, חזה, חזז, חטב, חטט, חטף, חסל, חסם, חסף, חצב, חצה, חצץ, חצר in the.
The root הם expresses the sound of humming, which is made with the mouth closed (μύω); hence הָמַם, הוּם, הָמָה, (נָאַם) נָהַם Arab. hámhama, to buzz, to hum, to snarl, &c.
As developments from the root רע cf. the stems רָעַד, רָעַל, רָעַם, רָעַע, רָעַץ, רָעַשׁ. Not less numerous are the developments of the root בר (פר, פל) and many others.
Closer investigation of the subject suggests the following observations:
30i (a) These roots are mere abstractions from stems in actual use, and are themselves not used. They represent rather the hidden germs (semina) of the stems which appear in the language. Yet these stems are sometimes so short as to consist simply of the elements of the root itself, e.g. תַּם to be finished, קַל light. The ascertaining of the root and its meaning, although in many ways very difficult and hazardous, is of great lexicographical importance. It is a wholly different and much contested question whether there ever was a period in the development of the Semitic languages when purely biliteral roots, either isolated and invariable or combined with inflexions, served for the communication of thought. In such a case it would have to be admitted, that the language at first expressed extremely few elementary ideas, which were only gradually extended by additions to denote more delicate shades of meaning. At all events this process of transformation would belong to a period of the language which is entirely outside our range. At the most only the gradual multiplication of stems by means of phonetic change (see below) can be historically proved.
30k (b) Many of these monosyllabic words are clearly imitations of sounds, and sometimes coincide with roots of a similar meaning in the Indo-Germanic family of languages (§1h). Of other roots there is definite evidence that Semitic linguistic consciousness regarded them as onomatopoetic, whilst the Indo-Germanie instinct fails to recognize in them any imitation of sound.
30l (c) Stems with the harder, stronger consonants are in general (§6r) to be regarded as the older, from which a number of later stems probably arose through softening of the consonants; cf. פזר and בזר, צחק and שׂחק, צעק and זעק, עלץ and עלז, עלס; רקק and רכךְ, and the almost consistent change of initial ו to י. In other instances, however, the harder stems have only been adopted at a later period from Aramaic, e.g. טעה, Hebr. תעה. Finally in many cases the harder and softer stems may have been in use together from the first, thus often distinguishing, by a kind of sound-painting, the intensive action from the less intensive; see above קצץ to cut, גזז to shear, &c.
30m (d) When two consonants are united to form a root they are usually either both emphatic or both middle-hard or both soft, e.g. קץ, קט, כס, גז, גד never כץ, גץ, גט, גס, קז. Within (triliteral) stems the first and second consonants are never identical. The apparent exceptions are either due to reduplication of the root, e.g. דדח (ψ 425, Is 3815), Arabic דאדא, or result from other causes, cf. e.g. בבּה in the Lexicon. The first and third consonants are very seldom identical except in what are called concave stems (with middle ו or י), e.g. נון, צוץ; note, however, נגן, נתן, שׁמשׁ, שׁרשׁ, and on עלע Jb 3930 see §55f. The second and third consonants on the other hand are very frequently identical, see § 67.
30n (e) The softening mentioned under l is sometimes so great that strong consonants, especially in the middle of the stem, actually pass into vowels: cf. §19o, and עֲזָאזֵל Lv 168 ff. if is for עֲזַלְזֵל.
30o (f) Some of the cases in which triliteral stems cannot with certainty be traced back to a biliteral root, may be due to a combination of two roots—a simple method of forming expressions to correspond to more complex ideas.
30p 3. Stems of four, or even (in the case of nouns) of five consonants are secondary formations. They arise from an extension of the triliteral stem: (a) by addition of a fourth stem-consonant; (b) in some eases perhaps by composition and contraction of two triliteral stems, by which means even quinquiliterals are produced. Stems which have arisen from reduplication of the biliteral root, or from the mere repetition of one or two of the three original stem-consonants, e.g. כִּלְכֵּל from כול or כיל, סְחַרְחַר from סחר, are usually not regarded as quadriliterals or quinqueliterals, but as conjugational forms (§ 55); so also the few words which are formed with the prefix שׁ, as שַׁלְהֶ֫בֶת flame from לָהַב, correspond to the Aramaic conjugation Šaphʿēl, שַׁלְהֵב.
30q Rem. on (a). The letters r and l, especially, are inserted between the first and second radicals, e.g. כָּסַם, כִּרְסֵם to eat up; שַׁרְבִּיט = שֵׁ֫בֶט sceptre (this insertion of an r is especially frequent in Aramaic); זַלְעָפָה hot wind from זָעַף to be hot. Cf. Aram. עַרְגֵּל to roll, expanded from עַגֵּל (conjugation Paʿēl, corresponding to the Hebrew Piʿēl). In Latin there is a similar expansion of fid, scid, tud, jug into findo, scindo, tundo, jungo. At the end of words the commonest expansion is by means of ל and ן, e.g. גַּרְזֶן axe, כַּרְמֶל garden-land (from כֶּ֫רֶם), גִּבְעֹל corolla (גָּבִיעַ cup); cf. § 85, xi.
30r Rem. on (b). Forms such as צְפַרְדֵּעַ frog, חֲבַצֶּ֫לֶת meadow-saffron, צַלְמָוֶת shadow of death, were long regarded as compounds, though the explanation of them all was uncertain. Many words of this class, which earlier scholars attempted to explain from Hebrew sources, have since proved to be loan-words (§1i), and consequently need no longer be taken into account.
30s 4. A special class of formations, distinct from the fully developed stems of three or four consonants, are (a) the Interjections (§ 105), which, as being direct imitations of natural sounds, are independent of the ordinary formative laws; (b) the Pronouns. Whether these are to be regarded as the mutilated remains of early developed stems, or as relics of a period of language when the formation of stems followed different laws, must remain undecided. At all events, the many peculiarities of their formation require special treatment (§32ff.). On the other hand, most of the particles (adverbs, prepositions, conjunctions) seem to have arisen in Hebrew from fully developed stems, although in many instances, in consequence of extreme shortening, the underlying stem is no longer recognizable (see §99ff.).
- On the questions discussed here compare the bibliography at the head of § 79.
- Cf. Philippi, ‘Der Grundstamm des starken Verbums,’ in Morgenländische Forschungen, Leipz. 1875, pp. 69–106.
- That all triliteral stems are derived from biliterals (as König, Lehrg. ii. 1, 370; M. Lambert in Studies in honour of A. Kohut, Berl. 1897, p. 354 ff.) cannot be definitely proved.
- Cf. the interesting examination of the Semitic roots QR, KR, XR, by P. Haupt in the Amer. Journ. of Sem. Lang., xxiii (1907), p. 241 ff.
- Consonants which are not found together in roots and stems are called incompatible. They are chiefly consonants belonging to the same class, e.g. גכ, גק, כק, דט, תט, בף, מף, זד, זס, זץ, צס, אע, חע, &c., or in the reverse order.
- In Hebrew they are comparatively rare, but more numerous in the other Semitic languages, especially in Ethiopic.
- So expressly Nöldeke in ZAW. 1897, p. 183 ff.; but most probably it is to be read צַלְמוּת darkness from the stem צלם [Arab. ẕalima, to be dark].
- Cf. Hupfeld, ‘System der semitischen Demonstrativbildung,’ in the Ztschr. f. d. Kunde des Morgenl., vol. ii. pp. 124 ff., 427 ff.