Hebrew Language
In large measure, the OT Hebrew must be self-explanatory. However, the Ugaritic Ras Shamra Tablets shed much light on the meaning of the Hebrew Bible, and since the structure and vocabulary were so very similar in the various Semitic tongues, much cognate language help is available for the understanding of the language of the Israelites. The Greek translation of the OT, the LXX, is also of much value in interpretative study of biblical Hebrew.
Though Aramaic is itself a very ancient language and the presence of “Aramaisms” in the OT often indicates an early rather than a late date for the passages in which they occur, from the time of the Exile onward Hebrew was spoken less and less and correspondingly the use of Aramaic flourished. Some of the Dead Sea Scrolls were written in Hebrew. Hebrew was also the vehicle for the writing of such Jewish religious literature as the Mishna and the Midrashim in the early part of the Christian era and in medieval times for biblical commentaries and philosophical and literary works. In modern Israel, Hebrew has again become a living tongue.
The historical origins of the language are somewhat obscure but go back beyond 2000 b.c. The OT literature, written over a period of more than a thousand years, reveals a minimum of stylistic changes, though loan words and new ways of expression became more or less noticeable with the passing of years, especially after the Exile. It is also true that at a given time dialectical differences existed, a fact attested by the narrative in Judg.12.1-Judg.12.15, in which Ephraimites were unable to pronounce the “sh” of their neighbors to the south.
With its short sentences and simple coordinating conjunctions, ancient Hebrew lent itself well to the vivid expression of events. These features, together with parallelism and rhythm and special meanings and constructions, made Hebrew poetry, as found in the Psalms and to a large extent in the Prophets, most expressive and strikingly effective.
HEBREW LANGUAGE. With the exception of Daniel 2-6 and portions of Ezra 4-7 (which were written in Aram.) the OT was composed in Heb., and has been preserved in its Massoretic form, equipped with vowel-points and accents, which date from the 8th or 9th cent. a.d. There are also portions of the Scripture in its consonantal form (i.e., in consonants written without vowel signs) dating from the 2nd cent. b.c. (the Dead Sea Isaiah Scroll) or perhaps even a cent. or two earlier (in the case of Pentateuch fragments written in Paleo-Heb. script). This means, of course, that all of the OT, whether ancient or modern, depend for their validity upon the faithfulness and adequacy of their rendering of the Heb. original. It also means that no interpretation or exegesis of the OT is valid apart from an adequate mastery of Heb. lexicography, grammar, syntax and rhetoric. To the Lord Jesus Christ and the authors of the NT, the OT, esp. in its Heb. original form, constituted Holy Scripture. Many of the theological terms employed by the Gr.-speaking authors of the NT were transferred from specific Heb. terms in the OT, and therefore rest upon them for valid interpretation.
The position of Hebrew in the Semitic language group.
Broadly speaking, the Sem. family of languages is divided into E Sem. (Akkad. with its sub-dialects of Babylonian and Assyrian), S Sem. (comprising N Arabic, S Arabic, and Ethiopic), and NW Sem. (comprising Canaanite, in its many varieties, and Aram. with its later offshoot, Syrian). Canaanite may have included Amorite (which has been preserved largely in the form of personal names), but certainly Sinaitic, Ugaritic, Phoen. with its Punic derivatives, Moabite and Heb.
Sinaitic inscrs., written in a type of alphabetic script resembling hieroglyphics, have been found not only in the turquoise mines of Serabit el-Khadim in the Sinai peninsula but also in Gezer, Lachish, Beth-Shemesh, Megiddo, and even Byblos in Phoenicia. They have been dated from the 18th to the 15th centuries b.c., and represent a type of S Canaanite which still preserved certain phonemes which coalesced into other phonemes in the Heb. of the OT (i.e., dh which coalesced with z, ḫ which coalesced with ḥ, and ə or ghayin which coalesced with ’ayin in Biblical Heb.).
Ugaritic, a dialect of N Canaanite spoken in Ras Shamra (ancient Ugarit) S of Antioch, was written in a thirty-character cuneiform alphabet, and it preserved most of the original Sem. phonemes which were assimilated to other phonemes in Biblical Heb. and Phoen. (i.e. th, which coalesced with sh; and z which coalesced with ṩ, in addition to the three in Sinaitic just mentioned above). A large quantity of clay tablets composed in this language was discovered at Ras Shamra, comprising letters, business documents, and a great deal of religious lit., esp. in the form of epic poetry relating the adventures of their gods and legendary heroes.
Phoenician was the dialect of Canaanite spoken in the great coastal cities of Tyre, Sidon, Berytus, Tripolis and Byblos, and in all of the colonies settled by Phoen. emigrants in Cyprus, Sardinia, Sicily and N Africa (in which latter area regional peculiarities developed resulting in Punic). It was apparently the Phoen. alphabet which was adopted by all the Canaanite-speaking peoples, including Hebrews and Moabites, and also by the Aramaeans as well, at least by the 11th cent. if not earlier. By this period the above mentioned consonant shifts had presumably taken place: i.e., dh had fallen together with z, rough ḫ with ḥ, ghayin with ’ayin, th with sh and z with ṩ (saddhe). Yet the ghayin sound may have been retained in actual speech, for even by LXX times this distinction was preserved in the Gr. transcription of ’-m-r-h as “Gomorrha” and of ’-z-h as “Gaza.” Phoenician developed characteristics which set it apart from Heb. Notably, the so-called “Canaanite shift” (original long a, still preserved as a in Ugaritic, already by Amarna times had shifted to long o) operated with a’s which became long under accent, even in the case of the secondary shift to the ultima in verb forms. Thus, Phoen. yaton (“he gave”), yatno (“she gave”), as contrasted with Heb. nātan and nātena. The causative stem in the Phoen. verb was a yiphil in place of the Heb. hiphil. Phoenician inscrs. date from the epitaph of Ahiram of Byblos (dated from 13th cent. to 11th cent. by various experts) to Punic documents of the 3rd cent. a.d. From Phoenicia itself the bulk of the material consists of royal stelae dating from 5th to 2nd cent. b.c.; from Cyprus mostly from 4th cent.; a few come from Attica and Egypt from about the same period; slightly later are the inscrs. from Sicil y, Sardinia (esp. Nora), Malta and Marseilles. The Punic material dates from the 4th cent. to 2nd or 3rd cent. a.d.
Moabite was really a dialect of Heb., and differed from Biblical Judaean Heb. only in the use of final -t for final -h in the fem. sing. absolute ending. The knowledge of Moabite is derived largely from the Mesha Stone, dating from about 840 b.c.; other Moabite inscrs. are fragmentary and hardly legible.
Concerning Israel’s other neighbors, the Edomites, Ammonites and Philistines, the surviving epigraphic evidence indicates that they too used Canaanite dialects very closely related to Heb., and the OT historical narrative suggests that they communicated with each other without any significant linguistic barrier. This does not mean that the original Philistine settlers from Crete were Canaanite-speaking; they may have used an Achaean dialect of Gr. similar to that of Cretan Linear B. Yet the apparently Sem. character of Cretan Linear A (somewhat Phoen. in its orientation, according to Cyrus Gordon) would suggest that they had some acquaintance with Canaanite even before their 2nd millennium migrations to S Pal.
Bibliography
For NW Sem. languages discussed above the following works will prove useful.
Comparative Semitics: M. Lidzbarski, Ephemeris für semitische Epigraphik, 3 vols. (1902-1915); W. B. Stevenson, Grammar of Palestinian Jewish Aramaic (1924); L. G. Gray, Introduction to Semitic Comparative Linguistics (1934); Z. S. Harris, A Grammar of the Phoenician Language (1936); Z. S. Harris, The Development of the Canaanite Dialects (1939); T. C. Vriesen and J. H. Hospers, Palestine Inscriptions (1951).
Northwest Semitic Languages: C. H. Gordon, Ugaritic Manual (1955); S. Moscati, A. Spitaler, E. Ullendorff and W. von Soden, An Introduction to the Comparative Grammar of the Semitic Languages (1964); W. F. Albright, The Proto-Sinaitic Inscriptions and their Decipherment (1966); F. Rosenthal, A Grammar of Biblical Aramaic (1968).
We now wish to say something about the lexical relationship of Heb. to other Sem. languages. It is only to be expected that the other Canaanite dialects are of great importance in illuminating the grammatical and lexical usages of Biblical Heb., although unfortunately the amount of lit. in these languages is regrettably meager, with the exception of Ugaritic. The decipherment and analysis of this N Canaanite tongue, totally unknown to modern scholarship until 1929, has proven to be of major importance in establishing the currency of words and phrases back in the Mosaic period which formerly were suspected by many OT critics of being textual errors in the Heb. Scriptures. New meanings for rare Biblical words have been made available by their occurrence in Ugaritic lit., and many obscurities thereby have been alleviated or cleared up altogether. Another important source of information concerning the meaning of obscure Biblical terms is the considerable lit. preserved in Aram. To be sure, there is a rather limited amount of this from the OT period, largely consisting of funerary inscrs., biographical epitaphs, governmental decrees, and a few epistles. Most of the Aram. from Intertestamental and NT times consists of the writings of Jews who were rooted in the OT tradition. The Genesis Apocryphon, discovered in Qumran Cave One, was written in the 1st cent. b.c. or earlier (in a form of Aram. separated by centuries of development from the Aram. of Daniel and Ezra); the Targum of Onkelos on the Pentateuch can be dated no earlier than a.d. 200, and the Targum of Jonathan on the prophets somewhat later than that. The Aram. of the Midrash and Talmud is later still. The Syr. Peshitta (written in an eastern dialect of Aram., and in a different form of alphabet from the “Square Hebrew” of Biblical and Post-Biblical Aram.) can hardly be older than the early 3rd cent. a.d.
Outside of the NW Sem. group, important sources of lexical data come from Akkad. (Assyro-Babylonian) lit., all of which comes from OT times or even earlier, and some of which deals with matters discussed in the Heb. Scriptures. Beginning with the Old Akkad. inscrs. of Sargon of Agade (c. 2200 b.c.), the First Dynasty of Babylon with its Code of Hammurabi (c. 1700), the Mari Tablets from the same period, the Nuzi Tablets from 1500, the Old Assyrian letters and commercial documents from the Assyrian merchant colony in Cappadocia, the royal inscrs., official annals, and business documents from the Assyrian Empire period (850-612 b.c.), and the Neo-Babylonian inscrs. and annals from the 6th cent., all of these furnish a large body of linguistic data from which to receive help in solving lexical problems.
As far as S Sem. is concerned, there is a modest number of dedicatory, boundary-marker and funerary inscrs. in the various dialects (Sabaean, Minaean, Qatabanian, Hadramautian) of Old S Arab. Northern Arab. material prior to Mohammed consists largely of short graffiti (Thamudic, Lihyanic, Safahitic) dating from 1st cent. b.c. to 4th cent. a.d. Beginning with the Qur’an and the early Islamic period a tremendous wealth of vocabulary has been preserved in a Sem. dialect so conservative in character as to preserve almost unchanged the phonology and morphology of Primitive Sem. Literary Arabic is of major value in supplying the most likely vowel patterns for the vowelless scripts of Ugaritic, Phoenician, Old Aramaic, and even of the earliest (Mosaic) form of Biblical Hebrew, which may still have retained short vowel endings on its nouns and verbs during the 2nd millennium b.c. Otherwise, one has to rely on the transcriptions of Canaanite names contained in Middle Kingdom and New Kingdom Egyp. inscrs., and about thirty-six different Canaanite words (the so-called “Canaanite glosses”) inserted into the Akkad. texts of the Tell el-Amarna Letters (c. 1420-1370 b.c.) in cuneiform syllabic spelling, clearly demonstrating that the short vowel endings were still in use back in that period.
The chief extra-Biblical Heb. documents from the OT period include the following: (1) a schoolboy’s exercise on a clay tablet listing the agricultural significance of the successive months of the year; dated at about 925 b.c. and discovered at Gezer, it is known as the “Gezer Calendar”; (2) the “Siloam Inscription” dating from about 705 b.c. in the reign of Hezekiah, and describing the completion of the tunnel cut through from the pool of Siloam to a reservoir inside the city wall of Jerusalem; (3) the “Samaritan Ostraca” dating from the reign of Jeroboam II (c. 770 b.c.) and consisting of receipts for taxes paid to the royal treasury in the form of wine or oil; (4) the “Lachish Letters” dating from about 587 b.c., discovered at Tell ed-Duweir, and consisting largely of communications from the commander of a Jewish outpost to his superior officer in the city headquarters. Apart from these are a few fragmentary lists of names, one recently discovered at Arad, and another found in the Qumran area, dating back to a pre-exilic army outpost maintained for a time in this region. These documents have served to confirm the authenticity of the OT books as written in the period to which they purport to belong (thus, the Lachish Letters are strongly reminiscent of Jeremiah); but more important still, they indicate the way Heb. was spelled in the older period, and thus give a clue as to possible ways in which copyists’ errors may have arisen in the course of the transmission of the OT text.
The characteristics of the Hebrew language
Phonology.
The Heb. alphabet consists of twenty-two letters, all of which are consonants. With these consonants Heb. could express nearly every consonantal sound in Eng. except j and ch. But they also included sounds foreign to non-Sem. languages: (i) ḥet, a rough kh sound resembling the ch in Scotch “loch” or German “ach,” yet without actual vibration of the uvula (although some Heb. words may have had this rougher type of ḥet, the kh sound, judging from Arab. evidence): (ii) ṭet, a type of t formed by firmly placing the tip of the tongue on the alveolar ridge (behind the dental position of tau, the regular t) and building up pressure before releasing; (iii) ’ayin, a sort of grunt made by constricting the throat muscles behind the uvula; there was also a ghayin type, which preserved a sort of fricative g, somewhat resembling a very guttural French r; (iv) ṩadde, a type of s marked by firm placement of the tip of the tongue right behind the base of the teeth and expelling air in an intense hiss (the modern Israeli pronunciation of ṩadde as ts is only an approximation of this); (v) qof, a kind of k pronounced farther back in the throat and engaging the uvula. The distinction in pronunciation between the two letters for s, samekh and ṩin, seems to have consisted in the placement of the tip of the tongue beyond the alveolar ridge and close to the teeth, in the case of the in; whereas the samekh was as s made in the normal fashion. Yet the evidence on this last is not altogether clear, and other interpretations are possible. At any rate it seems clear that in the no rthern kingdom of the Ten Tribes the distinction was lost quite early; in Judah it faded away by the 6th cent. and (esp. under Aram. influence) samekh began to be substituted for śin in the spelling of words. It also should be observed that sometime after 500 b.c. the letters b, g, d, k, p and t became spirantized after vowel sounds (i.e., were pronounced v, gh, sonant th as in “this,” kh, f, and voiceless th, respectively) unless the consonant was doubled. Thus, kātab was pronounced kāthav, gādap was pronounced gāthaf (th as in “gather”); but kittēb was kittēv, not kiththēv. Note also that all Heb. words are accented on the ultima (last syllable) unless an accent mark is indicated over the penult (next to last syllable).
As for the Heb. vowels, they consisted of a short a (prob. pronounced like the a in “hand”) and a long a (like the a in “father”), a short e as in “met,” and a long e as in “café” (like French e-acute or German long e), a short i as in “pin” (or like ee as many Canadians pronounce “been”), and a long i as in “machine”; a short o (so-called qāmeṩ-ḥatūṭ) as in “bottle” or “obey,” and a long o as in “home”; a short u as in “put,” and a long u like the oo in “moon.” There also were two diphthongs: ay (pronounced like y in “sky”) and aw (pronounced like ou in “out”).
Morphology of nouns.
In the earliest period the Heb. nouns, as in other Sem. languages, possessed short vowel endings; i.e., the nominative case ended in -u, the genitive in -i, and the accusative in -a (the feminine noun generally ended in -atu, -ati, -ata). The masculine pl. was -îma for all three cases; the feminine pl. was -âtu or -âti. There was also a dual number, esp. for parts of the body existing in pairs (e.g. eyes, ears, wings, feet); the masc. dual ending was -ayin, and the fem. -âtayim. Already in the Mosaic period, however, the fem. long a’s had rounded to long o (the “Canaanite shift”); hence -ôtu in the pl. and -ôtayim in the dual. In the course of time the short vowel endings were dropped altogether, except that in some cases the accusative masc. -a was retained as a long a in order to express motion to which (so called “hē-directive,” since this was spelled with a final silent hē), or in rare cases, location at which (e.g. mizbeḥāh “on the altar,” from mizbēaḥ, “altar”). Unlike fem. nouns ending in accented -āh, the accent always fell on the penult in the case of a hē-directive. The genitive had to be expressed by word position after the case endings were lost; i.e. the possessor always followed right after the thing possessed. For example, “the word of the king” (originally dábar hammálki, from dábaru “word,” and málku “king”) was expressed debar hammélek. Note that the definite article, “the,” was expressed by ha- and the doubling of the first letter of the noun it self. The first member of this pair was put into what was called the “construct” state, which often differed slightly from its normal (or “absolute”) state. Partly because constructs lost their short vowel endings much earlier than did the absolutes (observe the example of dabar hammalki, coming from dabaru), and partly because the construct bore no principal accent (which always fell on the second noun) the constructs often differed slightly from their normal or absolute form. This was esp. true of feminine nouns. For example, the word for “queen” was originally malkatu, and this became malkāh in the OT period; but in the case of the construct the -t was retained; thus “the queen of Israel” was malkat Yis’ rā'ēl.
In the case of the masc. pl. the ending -êy was used for the construct; thus, from sûsîm, “horses” we get sûsêy hammélek for “the horses of the king.” Apart from these special cases the construct relationship was shown merely by the immediate juxtaposition of two nouns; thus, sûs malkāh meant “a horse of a queen,” even though in this case the construct form of sûs does not differ from its absolute form. One other important development ensued upon the loss of case endings, and that was the use of ’ēt as an indicator of the direct object, esp. if the noun object was definite (i.e. preceded by the definite article, ha). This was virtually obligatory in prose (although ’ēt is rarely used in poetry). Thus, “The king killed the horse” would be: Qāṭal hammélek ’ethassûs.
As for adjectives, they were treated exactly like nouns, and originally had the same case endings. If they were attributive, they had to follow their noun rather than preceding it, and they had to have the definite article if their noun did. Thus, “a good horse” was sûs ṭôb, but “the good horse” was hassûs haṭṭôb. On the other hand, if the adjective was predicate, it normally preceded the noun and was never definite: “The horse (is) good” would be tôb hassûs. Interestingly enough, Heb. was the first of the Canaanite dialects to develop the definite article, prob. because the definite article was used widely in Egypt. colloquial during the 18th dynasty before the Israelites migrated from Egypt to Pal.
Morphology of verbs.
Unlike the Indo-European languages, the Sem. tongues were not primarily interested in time values, but rather in mode of action. That is to say, they were not so much concerned with when the act took place as with its state of completion or non-completion. In general, the Heb. perfect tense viewed the action as complete, single or generic; the imperfect tense presented it as incomplete or repetitive or continual. As a practical matter, however, the perfect tense usually referred to past time, and the imperfect tense usually referred to present or future time. Thus, for purposes of convenience one may render the perf. 3rd sing. masc. of the verb “to write” (kātab) as “he wrote”; and its corresponding imperfect, yiktōb, as “he will write” or “he writes.” This always needs to be verified from the context; the surrounding setting of the imperfect verb may show that it referred to continual or incomplete action in past time; correspondingly, the perfect tense was used sometimes to express the immediate past, where in Eng. we would employ the present tense. The frequent formula, “Thus says the Lord” was Kōh ’āmar Yahweh; the verb ’āmar is in the perfect, suggesting that the prophet knows what Yahweh has said only because He has already told him. The perfect is used occasionally to express a future action which has been foretold by the Lord as part of His divine plan—and is therefore as certain of accomplishment as if it had already taken place. In each case the context will indicate quite clearly what the temporal locus is for each verb.
Basically the same antinomy between complete and incomplete action carries over into the “consecutive” or “conversive” tenses. Unlike other Sem. languages, Heb. developed a specialized use of the particle we, “and.” When the simple we preceded the impf. there was no change in its tense value; but if it was strengthened to wa- with the doubling of the first consonant of the verb form, then it had the effect of converting an impf. into a perfect in its function. Thus, yiktōb meant “he will write,” but wayyiktōb meant “and he wrote.” This seems to have originated from an old preterite tense like the preterite in Akkad., which had genuine time value (thus, Akkad. iprus, “he judged”). It never had the short vowel ending of the genuine impf.; the original conversive was yaktub, and the original impf. was yaktubu. Since both forms became yiktōb in the 1st millennium, the only factor showing past time was the connective particle wa-. In certain types of weak verb the difference is quite apparent; e.g., yāqûm means “he will arise” but wayyāgom, “and he arose” shows a shortened form of the verb resulting from its lack of an original short vowel ending. Related to this preterite conversive form is the use of the Arab. jussive after the negative lam. The jussive also lacked a short vowel ending; thus, lam yaktub in Arab. meant, “He did not write.” Then, by a principle of polarity (i.e. if a waw can change an impf. into a perfect, then a waw should change a perfect into an impf.) the perfect conversive (or “consecutive”) arose as a secondary development. Thus, kātab means “he wrote,” but wekātab means “and he will write” (or even, “that he might write”). At least ninety percent of the perfects preceded by waw in Biblical Heb. are converted; the remaining five percent or ten percent can be shown from the context to be unconverted.
The mention of the waw connective leads to a discussion of the remarkable versatility of the particle we. As a connective between clauses it may mean “and” (if the idea of clause B follows along with that of clause A), or “but” (if the idea of clause B stands in opposition to that of clause A), or “in order that” (if B explains the purpose of the action in A), or “while” (if B supplies the attendant circumstance of the action in A), or it may simply introduce the apodosis (the waw-consequential) of a preceding protasis in a conditional sentence. In other words, the waw simply operated as a plus sign, linking idea B together with idea A, and leaving it to the intelligence of the hearer to put them in the proper relationship to each other. Hebrew possessed other particles capable of expressing these coordinate and subordinate relationships more precisely (e.g. lema’an for “in order that,” ’abāl for “but”), but for the most part the simple waw was used for these various ideas.
Turning now from the indicative uses of the impf. we come to the subjunctive and jussive. After a waw of purpose or other purpose particle the impf. (or perfect conversive) might be used to express the subjunctive idea: “in order that he may/might write,” There was also a jussive form (yiktōb, “let him write”), which looked just like the indicative impf. in the strong verb, with the one exception of the hiphil jussive (the term hiphil will be explained shortly), where the impf. yaktīb (“he will cause to write”) contrasts with the jussive yaktēb (“let him cause to write”) and impf. conversive wayyaktēb (“and he caused to write”). But since this jussive, like the impf. conversive, went back to an original form which lacked short vowel endings (yaktub, as contrasted with the impf. yaktubu), distinctive jussive forms were maintained in certain types of weak verbs. Thus, from the verb bānāh (“build”) we have impf. yibneh (“he will build”), jussive yiben (“let him build”); from qūwm (“arise”) we have yāqūwm (“he will arise”), but yāqōm (“let him arise”).
The imperative of the verb consisted of the impf. form without the preformatives. Thus, tiktōb meant “thou wilt write” (impf.), but ketōb meant “write thou!” (imperative). The fem. sing. imperative ended in -ī, the masc. pl. in -ū, and the fem. pl. in -nāh. There were two types of infinitive: (a) the infinitive construct (ketōb, resembling the imperative), which could serve as a complement to a main verb, could be preceded by a preposition like le (“to”), be (“in,” but usually with the force of “when”), or ke (“as” or “as soon as”). It could also take suffix pronouns as its subjects or objects; (b) the infinitive absolute (kātōb), which could have no preposition before it or suffix pronoun after it, but served to emphasize or extend the action of the main verb which it accompanied, or else could take the place of the main verb altogether (in cases where the action was being emphasized, rather than the agent of the action).
Having discussed the function of the various tenses of the verb, we now come to an analysis of the structure of the perfect and impf. Semitic verbs were constructed on a consonantal skeleton consisting usually of three consonants, and indicating tense or stem by doubling one of the consonants or by the vowel pattern used with them. Given the pattern, one is able to identify the tense and stem of almost any verb, even though he does not know the root meaning of that verb. The pattern CāCaC (where “C” represents any radical consonant) always means the perfect active masc. 3rd person sing. of that verb (thus, kātab, “he wrote”; qāṭal, “he killed”; šāmar, “he guarded”). The pattern yiCCōC always means the impf. 3rd masc. sing. active; thus, yiktōb, “he will write”; yiqṭōl, “he will kill”; yišmōr, “he will guard.” The perfect tense inflected by the addition of sufformatives; thus the fem. 3rd person sing. ended in -âh, the 2nd masc. sing. in -tā, the 2nd fem. sing. in -t (originally -ti), the 1st pers. sing. in -tī, the 3rd person pl. in -ū, the 2nd pl. masc. in -tem, the 2nd pl. fem. in -ten, and the 1st person pl. in nū. (Thus, katábtā meant “thou wrotest”; kātebû meant “they wrote,” etc.)
The imperfect tense was inflected by preformatives and sufformatives as well. Thus, yi- for masc. 3rd sing., ti- for fem. 3rd sing. or for masc. 2nd sing. (these identical forms could be distinguished only in context), ti-CCeCī (i.e. the preformative ti- and the sufformative -ī) for the fem. 2nd sing.; ’e- for 1st pers. sing.; yi- preformative and -û (sometimes -ûn) for 3rd pl. masc.; ti- preformative and -nāh sufformative for 3rd pl. fem. (which happens to be identical for the 2nd pl. fem.); ti- preformative and -û sufformative for 2nd pl. masc.; ni- preformative for 1st pl. All of these preformatives are standard for the basic ground-stem or qal (“light”) conjugation, as it is technically called. If the root begins with a guttural (’ayin, hē, ḥet, or even ’alef) the vowel of the preformative preserves the original short a; e.g. ya’amōd (“he will stand”) instead of yi’mōd. Hollow verbs, such as qûm, “arise,” also preserve this a, but pretonically (i.e. in the syllable before the accent) lengthen it, since the preformative stands in an open syllable (i.e., a syllable that ends in a vowel). Hence, “he will arise” is yāqûm. If the theme vowel of the imperfect qal happens to be a rather than ō (from an original u), then the preformative vowel was originally i, and this is preserved as e before roots beginning with a guttural; thus, “he will be strong” is yeḥezaq (rather than yaḥazaq) from ḥāzaq (“be strong”).
In common with the other Sem. languages, Heb. verbs possessed six derived stems (other languages possessed even more, like Arab. with its nine or more), which modified the root idea in various ways. These were: (1) the niphal (a term derived from the classical model verb, pā’al, “to make”—thus, the niphal of pā’al was nip’al) was a conjugation or stem which put the root into a reflexive or passive meaning. Thus, from kātab (“write”) the niphal form, niktab, meant either “he wrote himself,” or “he was written” (the passive meaning was much more common than the reflexive in this niphal stem). (2) The piel (pi’ēl, from pā’al) involved the doubling of the middle consonant, and it either intensified the action (thus, šibbēr, “shatter,” from šabār, “break”), or else it rendered the action transitive or causative. From qādaš, “be holy” the piel qiddēs meant “sanctify.” (Some verbs nearly always occur in the piel without any qal in use; e.g., dibbēr, “speak.”) (3) The passive of piel was pual; thus, šubbar meant “he was shattered,” and quddaš meant “he was sanctified.” (4) The reflexive of piel was hitpāēl; thus, hitqaddēš meant “he sanctified himself.” (5) The stem hiphil (hip’īl from pā’al) was primarily causative; thus, hiktîb, “he caused to write,” from kātab (“he wrote”). Some hiphils, however, were characteristic, i.e. they showed some noun in action. Thus, from the noun ’ōzen, “ear,” comes the hi phil he’ezîn (note how hi- modifies to he- before a guttural), meaning “to use the ear,” i.e. “to give ear” or “listen.” (6) The passive form of hiphil was hophal (hop’al); e.g. hoktab meant “he was caused to write.” Each verb might theoretically appear in seven stems (including the qal); an example of this would be qadaš: (i) qal: qādaš, “he was holy”; (ii) niphal: niqdaš, “he was hallowed”; (iii) piel: qiddēš, “he sanctified”; (iv) pual: quddaš, “he was sanctified”; (v) hithpael: hitqaddēš, “he sanctified himself”; (vi) hiphil: hiqdîš, “he made holy”; (vii) hophal: hoqdaš, “he was made holy.” In actuality, however, few verbs appear in all seven stems in Biblical Heb.; some of them occur in only one or two.
Space will not permit a full explanation of the various types of “weak” verbs, but in general, if any of the three radicals were weak, they were subject to modification or even disappearance under certain conditions, and they often resulted in alteration of the accompanying vowels. Thus, we have already seen that an initial guttural in the root would prevent the attenuation of an original short a to short i (thus, ya’amōd instead of yi’mōd) and would compel an original i to become e (thus, yeḥezaq instead of yiḥzaq). Similar changes were occasioned by a guttural in whichever position it was present in the root, whether first radical, second or third. Then again, if the first radical was a nūn or n, this became assimilated to the middle radical whenever no vowel sound separated them; thus, yippōl resulted from napal (“he fell”)—the unassimilated form would be yinpōl. If the first radical was a yod or waw, this created a diphthong with the vowel of the preformative; thus, qal imperf. of yāšab (“dwell”) was yêšēb (from original yayšibu); its hiphil perfect was hôšîb (from original hawšiba). If the middle radical was waw or yod, the result was a “hollow” verb like qûm (“arise”) or bîn (“understand”). If it was the third radical that was waw or yod, this led to alterations such as bānāh (“he built”) from original banaya, or yibneh (“he will build”) from original yibnayu. Or again, if the third radical was the same as the second, the two would combine to form a doubled consonant whenever there was no vowel sound in between. Thus from sābab (“surround”) is derived yāsōbbû (or else, by a transfer of the doubling to the first radical, yissebû), “they will surround.”
It should be added that all seven stems possessed participles; thus, from qādaš (“be holy”) the qal was qōdēš (although this particular verb would use the adjective form, qādôs), the niphal was niqdās; the piel was meqaddēš; the pual was mequddāš; the hitpael was maqdīs; and the hophal was moqdāš. These participles could serve as simple adjectives (thus, hassōpēr hakkōtēb would be “the writing scribe”), or as relative clauses (the previous example could also be tr., “the scribe who was writing”), or as agential nouns like kōhēn, “priest” (from the root k-h-n, signifying “serve as a priest”), or as the predicate of a clause after a noun or pronoun subject. This last construction was known as a present periphrastic, and was used to convey either continuous, durative action (e.g. hassōper kōtēb, “the scribe is writing”) or else imminent action (e.g. Yahweh nōtēn lākem could mean, “Yahweh is about to give to you”—from nātan, “give”).
Pronouns.
The independent pronouns were: ’anî or ’ānōkî (I), ’attāh (thou, masc.), ’att (thou, fem.), hû (he), hî (she), ’anaḥnû (we), ’attem (ye, masc.), ’atten (ye, fem.), hēm (they, masc.), hēnnâh (they, fem.). The possessive pronouns were expressed by a set of enclitics: -īy (my), -kā (thy, masc.), -ēk (thy, fem.), -ōw (his), -āh (her), -ēnû (our), -kem (your, masc.), -ken (your, fem.), -ām (their, masc.), -ān (their, fem.). These were written as part of the same word cluster as their noun; thus, debārekā, “thy word” (from dābār, “word”). They were added to nouns in their construct form, both in the sing. and the pl. Thus the fem. nouns would show an -at ending before the possessive pronoun was added; or, if pl., the ending -ôt. Thus “thy queen” was malkātekā, “your queens” malkôtêkem (the -ôt construct ending was followed by the masc. pl. construct ending, -êy, before the pronoun was added). The masc. pl. nouns produced some changes in the pronoun because of this, -êy; thus, “his kings” was melākā&sup(y);w, with a silent yod inserted in order to make it clear to the reader of an unpointed text that the noun was pl., not sing. “Her kings” was melākeyhā. It should be observed that with the inseparable (proclitic) prepositions prefixed to nouns (be “in,” le “to,” ke “like, as”), there resulted in a sizable cluster, esp. if a possessive pronoun was appended. From ’ereṩ “land” (originally ’arṩu) the combination be’arṩenū meant “in our land”; “to thy words” was lidebāreykā. As for the demonstrative pronouns, “this” (masc.) was zeh, “this” (fem.) was zō’t, and “these” was ’ēlleh. To express “that” or “those” the independent third person pronouns were used, preceded by the article. Thus, “that king” was hammelék hahû, “that queen” was hammalkāh hahî, and “those prophets” (from nābî' “prophet”) was hannebî'îm hahēm. The interrogative pronouns were mîy (“who” personal) and mâh (“what” impersonal). The relative pronoun was ’ašer, more rarely še (esp. in poetry), although often the relative clause was expressed by a participle with the definite article (thus, ’elōhîm habbōrē’ ’et-hā'āreṩ means “God, who created the earth”), or else it was omitted altogether (cf. the Eng.: “Women I have known”).
Word order.
The normal word order of Heb. prose was: (1) the verb, (2) pronoun object or indirect object, (3) noun subject, (4) noun object. Thus, “The king gave me the book” would be: Nātan lî hammélek ’ethassēper. Any deviation from this order indicated some special emphasis. Thus, Lî nātan hammelek ’et-hassēper would mean, “It was to me that the king gave the book.” Correspondingly, ’Et-hassēper nātan lî hammélek would mean, “It was the book the king gave me.” Or, Hammélek nātan îl ’et-hassēper would mean, “It was the king who gave me the book.” In poetry the same general word order prevailed, but emphasis often influenced deviations from the verb-subject-object pattern. Introductory temporal or causal phrases would of course precede the verb; likewise such conjunctions as kî (“when, because, if”), ’im (“if”), or the relative pronoun, etc. Special emphasis also was achieved by the use of the independent pronoun with an inflected verb. Thus ’attâh kātabtā would mean, “It was thou who wrotest”; ’anī ’adabbēr means, “I myself will speak.”
Loanwords in Hebrew.
General characteristics of Hebrew style.
From the foregoing description of the morphology of Biblical Heb. it will be apparent that it lacked precision in tenses such as characterize Gr., Latin and the Indo-European languages generally. It possessed remarkable power as a medium of emotional expression and prophetic eloquence. Because of its ability to compress so much thought-content into few syllables, it possessed a dynamic, penetrative force capable of stirring the hearer to the depths of his being. It may have lacked the ability of Gr. to express fine shades of meaning, and many of its key terms may be difficult to define exactly, but it possessed a remarkable ability to soar to sublimer heights in its beautiful parallelistic poetry or exalted prophetic passages than other languages could hope to attain. In sound it was unusually sonorous, more so than the other Sem. languages, because of the frequency of pretonic long syllables (a feature peculiar to Heb.). No tr. into another tongue can do justice to the emotional intensity and rugged beauty of its verse (perhaps Ger. comes the closest to this ideal), or duplicate its shifting moods of exaltation and pathos. Its theological vocabulary requires careful study in depth, if the overtones, undertones and connotations are to be grasped, for these are beyond the power of even the most skillful tr. to convey with any real adequacy. Truly this was a linguistic medium ideally adapted for the proclamation of God’s message to man, so as to reach his intellect, his emotions and his will.
Bibliography
The standard Heb. grammar in Eng. is Gesenius-Kautzsch-Cowley: Gesenius’ Hebrew Grammar (2d ed.), 1910 (Oxford). Perhaps the most comprehensive grammar in Ger. is Bauer and Leander, “Historische Grammatik der hebräischen Sprache des Alten Testaments (1918-1922). In French one of the best is P. Joüon, Grammaire de l’hébreu biblique (1923). For a systematic treatment of the verb the basic treatise is still S. R. Driver, A Treatise on the Use of the Tenses in Hebrew (1881). As for dictionaries, the best lexicon in English is still Brown-Driver-Briggs, A Hebrew and English Lexicon of the O. T. (1906). In German, F. Buhl, Wilhelm Gesenius’ hebräisches und aramäisches Handwörterbuch über das Alte Testament (repr. 1949). Koehler-Baumgartner, Lexicon in Veteris Testamenti Libros (Brill, 1951) tr. into both Ger. and Eng., but omits many of the rare words through the device of amending them to commoner ones. The standard scholarly text of the Heb. Scriptures is R. Kittel, Biblia Hebraica (7th ed.), Stuttgart, copyright 1937. A fine companion volume to this is Ernst Würthwein, The Text of the O. T. (Ackroyd tr. 1957). For most comprehensive information on the Qumran discoveries of Heb. MSS: M. Burrows: The Dead Sea Scrolls (1955) and, More Light on the Dead Sea Scrolls (1958).