The Hebrew language, or more accurately Biblical Hebrew, in the main the language of the Hebrew Bible is also called Early Hebrew, in contradistinction to the language's later stages (Rabbinic, Medieval, and Modern). It belongs, with, for example, Moabite and Phoenician, to the Canaanite family of Northwest Semitic languages (to which Ugaritic and Aramaic also belong). [See Phoenician-Punic; Ugaritic; Aramaic.]

Biblical Hebrew.

As a Semitic language, consonants in Biblical Hebrew almost exclusively constitute its roots, while its vowels modify them, through the formation of verbal and nominal themes. As a West Semitic language, its suffix conjugation (e.g., kāṯaḇtī, “I wrote”) has become a veritable verbal form denoting the past, rather than a state (as in East Semitic). By not expanding the use of broken plurals and of the conative verbal theme fā῾ala, as have the Southwest Semitic tongues (e.g., Arabic, Ethiopic); and by shifting w in initial position to y, Hebrew pertains to the Northwest Semitic languages. [See Arabic; Ethiopic.] By exhibiting for example, the consonantal shifts ā > ō, ḏ > z, ṯ > š, ḍ/ > and showing the vowels i–i, rather than a–a, in the active derived verbal themes; and by forming many participles from the stem of the suffix conjugation (qām, zāqēn, nišmār) and themes like rōmēm, “he exalted,” rather than ⋆ riyyēm, in the hollow and geminate verbs, it belongs to the Canaanite family (although the last two features occur in Ugaritic as well).


Biblical Hebrew is mainly known from the predominantly Hebrew parts of the Bible (only Dn. 2:4–7:28, Ezr. 4:8–6:18, further Jer. 10:11, and two words in Gn. 31:47 are in Aramaic). It is transmitted in a consonantal script of which the characters are historically Aramaic, adopted for the (also consonantal) Old Hebrew alphabet following the destruction of the First Temple. Inscriptions written during the period of the First Temple were in the Old Hebrew script and, during the Second Temple period, in Old Hebrew and Aramaic characters alternatively. Because they are written in a consonantal script, little precise information regarding the vocalization of the language can be elicited from inscriptions. Ancient transcriptions are also of uncertain value for determining vocalisms because of the pitfalls connected with transcription from one language to another. Nevertheless, they serve as an important corrective for the biblical material, transcriptions because they include vowel markings, and inscriptions because they provide new phonetic and morphological material. Moreover, inscriptions provide additions to the limited vocabulary of Biblical Hebrew.

Hebrew Language and Literature

HEBREW LANGUAGE AND LITERATURE. The Semitic background of the Roman alphabet. The letters tet and ṣade are excluded since they have no counterparts in the Roman alphabet.

view larger image

The transcriptions are, to a great extent, limited to the transliteration of proper nouns—for example, into Akkadian and Greek (as in the Septuagint [third century BCE onward]) or Latin (as by St. Jerome [342–420 CE]). Yet, Greek transcriptions are not restricted to proper nouns because Origen (185–254 CE) transliterated the Hebrew Bible into Greek letters and a few fragments of the transliteration of Psalms are preserved. [See Akkadian; Greek; Latin.]

The most important Old Hebrew inscriptions from the biblical period are the Gezer Calendar (c. tenth century BCE); the Kuntilet ῾Ajrud votive inscriptions (from the site on the Sinai border; late ninth century BCE); ostraca from Samaria (from the eighth [?] century BCE); the Murabba῾at palimpsest (Mur. 17), the oldest Hebrew papyrus (eighth–seventh centuries BCE); the Siloam tunnel inscription (c. 700 BCE); the Meṣad Ḥashavyahu ostraca; two small silver amulets found in Jerusalem in the Hinnom Valley, one of which contains the Priestly blessing also known from Numbers 6:24–26 (seventh century BCE); the Arad and Lachish letters (sixth century BCE); as well as a large number of inscribed seals. [See Gezer Calendar; Kuntillet ῾Ajrud; Samaria Ostraca; Murabba῾at; Siloam Tunnel Inscription; Jerusalem; Meṣad Ḥashavyahu Texts; Arad Inscriptions; Lachish Inscriptions; Seals.]

The biblical text itself is not easy to evaluate. It is made up of three distinct historical strata. In order of their antiquity and their importance, they are the consonantal text, the use of consonants as reading indicators (matres lectionis, “mothers of reading”; Heb., ῾immôt qĕrî'â), and the system of diacritical marks for vowels and cantillation. The twenty-two letters of the Hebrew alphabet were originally purely consonantal. Some were polyphonic—šin/śin—and distinguished even in modern traditional pronunciation as š/s, except according to the Samaritan tradition, and also presumably ῾ayn/ǵayn and ḥet/ḫet, respectively. This implies that the Hebrew speakers took over the alphabet from a dialect in which these consonants coincided, preferring to use these letters polyphonically, rather than invent new letters for the consonants not represented. A more original form of the alphabet, as reflected in Ugaritic, contained twenty-seven letters. (The Ugaritic alphabet also included three additional letters, added secondarily at the end of the alphabet.) The existence of twenty-nine consonantal phonemes is usually posited for Proto-Semitic. It stands to reason (and is also proved by deviating translations of the Septuagint) that the earliest stage of the biblical text was written in a purely consonantal script and that the use of matres lectiones came about through changes in pronunciation and subsequently through analogical usages and intentional reform. These matres lectionis—namely waw/yod marking o and u/e and i (originally ō and ū/ē and ī), respectively, and heh in word final only and ᾽alef only in words in which it had been used as a consonant—were increasingly utilized by the scribes of the Bible, yet they were a rather restricted aid for enabling the exact reading of the holy text. [See Writing and Writing Systems; Scribes and Scribal Techniques.]

It was for that reason that vowel and cantillation marks were introduced. This occurred at a rather late period (between 600 CE, the date of the final redaction of the Talmud, in which they are not yet mentioned, and the beginning of the tenth century, from which time dated manuscripts are known). The marks are, however, based on a much older tradition. The only vocalization and cantillation system used today by the Jewish community is the so-called Tiberian system. It is the most elaborate system and the only one completely preserved. Accordingly, it serves as the main basis for linguistic investigation. In principle, however, the other vocalization (and cantillation) systems are equally important: the Babylonian one (in which a and æ are not distinguished) and the so-called Palestinian system (a sub-system of which corresponds to modern Sephardi pronunciation used in academic teaching and in modern Israeli speech). The Tiberian vocalization differs from both the Babylonian and Palestinian by the shift of u to å, thus coinciding with å < ā. The Samaritan tradition is quite different.


Despite the multilayered character of the linguistic tradition, Biblical Hebrew though stretching over many hundred of years and stemming from different parts of ancient Palestine, reflects a surprisingly uniform language. This is the result of its being a standardized literary language, presumably emerging from Judea (Judah) in general and Jerusalem in particular, on the one hand, and to the later changes the that text (especially its vocalization) underwent, on the other. Nevertheless, a distinction must be made between poetry and prose, and then between early and classical poetry and between classical and late prose, without going into details and differentiating between various types of prose (e.g., historical, legal). Thus, poetry in general uses the longer (original) forms of the prepositions ᾽æle, “to”; ῾ale, “on”; and ῾ade, “until”; it restricts the use of the definite article, the relative pronoun, and the definite direct object marker; it sometimes adds the endings -i/o to nouns in construct (-i in other cases as well); it utilizes the pronoun -mo, “their/them”; and it uses construct forms preceding prepositional phrases. Early poetry has preserved the short prefix conjugation to mark the past (even when not following the “conversive” waw). For Late Biblical Hebrew, see below.


The Hebrew alphabet's twenty-two letters originally marked only consonants—῾b g d h w z ḥ ṭ y k l m n s ῾ p ṣ q r š/ś t. The adduced order of the letters is very old, reflected not only by internal evidence (i.e., by biblical acrostic), but also by the Ugaritic alphabet. The penultimate letter remained polyphonic, yet the lateral ś disappeared in pronunciation. It was replaced by s, identical with the fifteenth letter, so that the alphabet reflects the pronunciation of twenty-two letters only. Historically, however, it exhibits twenty-three phonemes (though in biblical orthography ś and s sometimes are interchanged). On the other hand, the letters b, g, d, k, p, t have allophones, for after vowels they are pronounced as spirants—ḇ, ḡ, ḏ, ḵ, p̄, ṯ. Yet, the automatic shift of these plosives (which are regularly marked in the Tiberian system by an internal point, the so-called dageš lene) to spirants (which are, in accurate manuscripts, marked by a superposed line, the so-called raphe) has been preserved in word-initial position only; elsewhere the spirant allophones are on the point of becoming phonemes (cf. låqaḥat, “you took” [fem. sg.], as against lāqaḥaṯ, “to take”; cf. also the very same word with alternating plosive and spirant pronunciation: qorbān, “offering,” as a rule with b; but see Ez. 40:43, qorḇān, with b).

The realization of the phonemic status of noninitial b, g, d, k, p, t is of utmost importance also for understanding the biblical vowel system: because the plosive/spirant pronunciation does not necessarily depend on being preceded by zero vowel or by ǝ respectively, ǝ must not be analyzed as a phoneme (see below). Between short vowels, h tends to disappear, as does aleph in certain positions, especially in syllable final; n preceding a consonant is almost invariably totally assimilated to it. The doubling of final consonants is given up (⋆ gabb > gaḇ; only final tt tends to be preserved, as ᾽att, “you” [fem. sg.]). This may also occur in a medial position preceding a consonant (tåsoḇnå, “they will turn” [fem.]), even when the originally following ǝ has only later become zero (as ⋆ wayyǝhī > wayhī, “and it was”). Final consonant clusters tend to be opened by an auxiliary vowel, usually æ, which, however, as a rule, does not acquire phonemic status. At a later phase, laryngeals/pharyngeals—', h, ḥ, ῾—became weakened. Therefore, not only did they (and r) lose their ability to be doubled, they also often trigger the insertion of ultrashort vowels (a, æ, å) when they are in word-medial position immediately followed by a consonant. These vowels also regularly replace ǝ after laryngeals/pharyngeals, which also tend to change i/u to a (i also to æ).

As to vowels, the Tiberian vowel system, with the exception of ultrashort vowels (a, æ, å), denotes only quality, rather than quantity. That is proven, inter alia, by the use of the same vowel sign to mark å both when it developed from long ā and short u. It also tends to mark allophones, not only, as mentioned above, b, g, d, k, p, t, whether plosive or spirant, in word-initial position, and auxiliary vowels opening a final consonantal cluster, but also when it is inserted automatically as a vocalic glide before final h, ḥ, ῾after vowels other than a/å (the so-called furtive pataḥ). It marks both the absence of a vowel and ǝ (according to Tiberian tradition, often pronounced a) by the same sign (šwa). Because the spirantization of b, g, d, k, p, t does not necessarily depend on a preceding vowel, including ǝ, and because ǝ often shifts to zero and vice versa, no phonemic status should be accorded to the opposition ǝ:zero.

Historically, the Hebrew vowels derive from the Semitic triads—short a:i:u, long ā:ī:ū. Whereas the number of the Semitic consonantal phonemes has been reduced in Hebrew, that of the vowels has increased. The original short vowels are preserved in closed, unstressed syllables. However, a is often “attenuated” to i (a rather late feature, in which the various vocalization systems differ), and u tends to shift to å, except when followed by a doubled consonant; a is also preserved in stressed closed syllables of verbs, as well as in nouns originally terminating in a doubled consonant, and in construct. Otherwise, a shifts to å, i to e, u to o. In stressed closed syllables, i tends to shift to a (sometimes also to æ)—the so-called Philippi's law, which occurred much later than generally assumed. In pretonic open syllables, a (and sometimes also i) shifts to å (and e, respectively; i, however in this position is often elided). Sometimes the consonant following the pretonic syllable is doubled, especially after u, and quite often this doubling spreads to the whole paradigm. Short open syllables two or four syllables before the stress are reduced. Long ī and ū are preserved (without length), whereas ā was originally kept in unstressed syllables, shifting to o in stressed ones (this distribution has, however, largely been altered by analogy). When unstressed, the diphthongs aw/ay were monophthongized to o/e, respectively—yet preserved, for example, preceding double w/y. In closed syllables bearing the main stress, they were split into two syllables, as ⋆ mawt > må:wæṯ, “death”; ⋆ bayt > ba:yiṯ, “house.”


Because the cantillation marks in the Tiberian system are, as a rule, on the stressed syllable, the Tiberian stress system is known with certainty. Synchronically, stress, as a rule, falls on the ultima, not rarely on the penult. A stage of general penult stress can be reconstructed historically because words stressed in pause on their ultima are generally forms that have elided final vowels; those stressed on their penult in pause, however, have preserved their final syllables. Adding the vowels elided to the words stressed on their ultima even in pause results in the great majority of Hebrew vocabulary stressed on the penult. In pause, the penult stress of words that did not elide their final vowel, was as a rule, preserved because the penult, even when originally short, was lengthened in pause. Long vowels, as a rule, preserved their penult stress (e.g., ⋆ kātā:bū, “they wrote,” with penult lengthening of the first syllable, according to the Tiberian system, in which no long vowels existed, kåṯå:ḇu). In context, however, the penult was often short, which caused the stress to move to the ultima. Because pretonic lengthening preceded the shift from penult to ultima, the first syllable of the context form ⋆ kāta:bū, “they wrote,” was preserved, being long. Pretonic lengthening had already ceased operating, however, and therefore the open short unstressed penult, now losing the stress, was reduced to ǝ: ⋆ kātǝbū: and, according to the Tiberian system, kåṯǝḇu:

This theory of general stress is of special importance for the proper understanding of Biblical morphology. Without this theory, chaos obtains; it enables discovering regularity almost everywhere, although it was often blurred by analogy. Thus, for example, the “conversive” and preceding the (generally short) prefix conjugation has the form wa, and the following consonant is doubled. Because, according to the theory of general penult stress, many forms of the short prefix conjugation, which consisted of two syllables only, were stressed on their first syllable, the a of wa was preserved by pretonic doubling.


In Hebrew in particular and in Semitic tongues in general, triradical consonantal structure prevails, and the main meaning is carried by the radicals. Vowels only add shades to it. Interjection and pronouns, because of their emotional character (which, to be sure, is to a large extent blurred in pronouns), deviate from this structure. Moreover, pronouns allow word composition, a feature alien to Semitic linguistic structure (with the exception of negations): for example, see hallåzæ, “this,” compounded of the demonstrative ha (which serves as the definite article as well)+the emphatic element +the demonstrative .

Although a substantive to which an attributive demonstrative is attached is automatically definite, it nevertheless is always also formally defined. When this is done by the definite article, it is attached to the demonstrative as well: ballay:1å hahu, “in that night”; yet, the more archaic form ballay:1å hu, in which the definite article is attached to the substantive only, is also attested.

The personal pronouns have basically two forms: the “nominatival” ones, serving as subject and predicate, are separate words. The “genitival” (denoting possession and governed by substantives and prepositions) and accusatival forms (denoting direct object and governed by [transitive] verbs) are enclitic and suffixed to the preceding governing word. The genitival and accusatival suffixes are identical, except for the first-person singular, “mine,” being stressed -i, “me” unstressed -ni. Most of these suffixes are preceded by “connecting” vowels, stemming from the short final vowels in which the preceding words are terminated; the absence of these vowels is the result of secondary development. It is only after the third-person feminine singular of the suffix conjugation, which originally terminated in a consonant (the suffix -aṯ), that the pronominal suffixes join the governing word directly, without a connecting vowel.

Tense system.

The Hebrew tense system, beside the imperative (in the second person only, in a form closely related to the short prefix conjugation), consists of four finite forms: the simple suffix and the conversive suffix conjugation, the simple prefix and the conversive (as a rule short) prefix conjugation. The prefix conversive forms are preceded by wa, “and.” It is a moot question whether the verbal forms originally denoted aspects or time. In classical biblical prose, at any rate, these forms seem to mark time: the difference between the simple suffix and the conversive (as a rule short) prefix conjugation, both referring to the past, and between the simple prefix and the conversive suffix conjugation, both referring to the future/present, respectively, depends on the syntactical environment only. As a rule, whenever the use of w, “and,” is possible, the “converted” forms are applied, in accordance with the demanded time; otherwise, the simple forms are used. This tense system, exhibiting two pairs of forms identical in meaning and differing in the syntactical environment only in which they occur, is quite unlooked-for. It was, to a great extent, therefore that Hans Bauer considered Biblical Hebrew to be a Mischsprache, a “mixed language.”

It is, however, with Gotthelt Bergsträsser (1918–1929, vol. 2, p. 3, n. 1), very difficult to imagine that two systems should be absorbed in their entirety into one language. Accordingly, it stands to reason that originally the simple suffix conjugation denoted state, as it does in Akkadian, whereas the past was marked by the short prefix conjugation (which in Akkadian and Arabic, as in Hebrew, combines the meanings of past and jussive). As a designation of the past, the short prefix conjugation was then superseded by the suffix conjugation. It was only preserved (except in archaic poetry) in this usage in the closed syntagma preceded by wa. Accordingly, from the historical point of view, this wa is not conversive but rather preservative. The “converted” suffix conjugation is, it seems, secondary. It is formed by the analogy of wa+prefix conjugation, perhaps also because wishes could be denoted by the past tense verbal form as well. In poetry, however, the use of tenses is quite free, presumably because of an intentional archaism as well as a pseudoar-chaism.

Besides the indicative, the prefix conjugation also denotes mood—some sort of volitive action: in the first person, the “cohortative,” a lengthened form (by the suffix -å), is utilized, whereas in the second and third persons, a short form, the “jussive,” is used. This jussive is basically identical with the short converted prefix conjugation after wa. It later differs from it in that the converted form often preserves the original penult stress.

Verbal themes.

It is in verbs that consonantal triliterality has evolved most completely. Although historical traces of biliteral roots can be discovered, synchronically the Hebrew verb has to be analyzed as triradical—no doubt a late development, the result of analogy. The impact of far-reaching analogy can also be felt in the very restricted number of verbal themes, as against a plethora of nominal themes. There are seven main verbal themes, all but one of which are named after the form of the third-person masculine singular of the suffix conjugation: the basic theme, without any affixation or doubling, called qal, “light”—that is, “light” as regards addition, without any addition; nif῾al, formed with prefixed n, originally only reflexive, later, after the original passive of the qal had fallen into desuetude, also passive. From the passive of the qal, the pu῾al, only some vestiges have been preserved. Even fewer have been preserved from its (reflexive-reciprocal) t-form (hiṯpå῾al) and the t-form of hif῾il, the hiṯaf῾al. Both the causative hif῾il and the doubled middle radical (intensive-factitive) pi῾῾el have (internal) passive forms, hof῾al (huf῾al)/pu῾῾al, respectively. The latter also has a (reflexive-reciprocal) t-form, the hiṯpa῾῾el. Accordingly, with the exclusion of nif῾al, the well-balanced system shown in table 1 can be reconstructed.

Among the nonfinite forms, the Hebrew verbal system possesses a participle (in the basic theme also a passive participle) that has both nominal and verbal rection. There are two “infinitives”: The so-called infinitive construct has the usual infinitive functions (but also that of a gerund); it is, as a rule, preceded, even when not used in a final sense, by the preposition , “to,” which may, however, when the infinitive clause does not have a final sense, be missing. The so-called infinitive absolute (so-called because it does not stand in construct and is not governed by prepositions) is a peculiar blend between verbal noun and verbal interjection. It is often used, in addition to its rare infinitive functions, for example, to emphasize the verbal action, as a rule preceding the main verb. It also serves as a modal adverb, presumably an archaic feature, because in Akkadian and Ugaritic it terminates in the adverbial ending -u. Finally, it may substitute for any finite verbal form, though the replacement of the imperative is particularly well attested.

In the suffix conjugation of the basic theme fientive få῾al tends to supersede stative få῾el and få῾ol. The original patterns of the prefix conjugation of the basic theme were fientive yaf῾il (which has been absorbed by hif῾il), yaf῾ul, and stative yif῾al. Yet, only vestiges of these patterns have been preserved, as yaḥaloš, “he weakens,” as against yæḥælaš, “he is weak”; yaqum, “he rises,” as against yeḇoš, “he is ashamed”; yåsoḇ, “he turns,” as against yeqal, “he is light.” In pi῾῾el the e of the suffix conjugation is mainly preserved in the third-person singular masculine in pause. In context, as well as in the other persons (except in the third-person singular feminine), the e, through Philippi's law, shifts to a.

Table 1. Verbal Themes. Obsolete forms are marked by an asterisk.

Basic Theme Double Theme Causative Theme
qal pi῾el hif῾il
Internal passive ⋆pu῾al pu῾῾al ho/uf῾al
ṯ-form (reflexive-reciprocal) ⋆hiṯpå῾al hiṯpa῾῾al ⋆hiṯaf῾el


Both triliteralism and the development of themes are less conspicuous in nouns than in verbs. There exists a set of biradical substantives with a fixed vowel. By their meanings they demonstrate that they belong to the oldest stratum of the language and are, even synchronically, best analyzed as biradical. Substantives are used in different states: in the absolute state (when standing alone), in construct (when closely attached to a following noun, the nomen rectum, historically a genitive, denoting relation of possession, etc.) and in the pronominal state (when attached to a genitival pronominal suffix). Two genders exist, the masculine, as a rule with zero ending, and the feminine, with -å (spelled -h in construct and pronominal state -at) and -t ending.


The cardinal numbers 3–10 reflect the very surprising common Semitic feature that, in opposition to the other noun classes, those with a zero ending modify feminine nouns; those with an -å ending (in construct -[a]ṯ) modify masculine ones. They precede or follow the counted noun in the absolute state, but they may precede it in construct—historically an archaic feature. This is the rule with definite nouns, as well as with nouns often counted, as yom, “day.” The “ten” in numbers 11–19 is ῾åśår for masculine, ῾æśre for feminine, spelled ῾śrh. The ordinal numbers have special forms only for the digits and ten built, as a rule, according to the theme pǝ῾ili. The exceptions are šišši, “sixth,” built secondarily off of šeš, “six,” and perhaps šeni, “second,” “first” is rišon, derived from roš, “head,” a relatively late form, as is customary in Semitic languages. The older usage of the cardinal number 1 appears in Genesis 1:5, yom῾æḥaḏ, “one day”=the first day.”

Clause formation.

It is in the domain of clause formation that Biblical Hebrew has best preserved the ancient Semitic character. In contradistinction to Arabic, it has not relinquished free sentence structure in favor of systemization. Yet, although it has lost, like Aramaic, case and mood endings, it has not been affected by a similar syntactic formlessness. The boundary lines between main and subordinate clauses, however, as well as between the coordination of sentences and subordination, is sometimes blurred. Thus, the main clause, when it follows the subordinate clauses, may be introduced by , “and.” In classical prose, main clauses, as a rule, open with a verb preceded by the conversive wa. Sentences describing background tend to open with the subject and may even be nominal. Yet, because the sentence following the background sentence may open again with wa+verb, either when it continues the background sentence or resumes the main action, it is often difficult to understand the exact sentence structure. Moreover, if the verbal clause does not open with the finite verb or with the subject (i.e., it begins with an object or an adverbial), the finite verb has to precede the subject (i.e., the obligatory word order is object/adverbial+finite verb+subject). An example in Genesis 1:1–3, bǝrešiṯ bårå' 'ælohim 'eṯ haššåma:yim wǝ'eṯ hå'å:ræṣ wǝhå'å:ræṣ håyṯå ṯo:hu wåḇo:hu…,wayyo:mær 'ælohim: “in the beginning God created the heaven and the earth [main action, opening with an adverbial], the earth being without form and void … [nominal clause serving as background], and God said [again main action; yet theoretically, it could have continued the background clause as well]. Quite frequently introduces the main clause in phrases like wayhi ki …, , “and it happened when…, then.…” After a jussive or an imperative, for example, the prefix conjugation following , by attraction to the preceding volitive, is transferred into the volitive mood (i.e., the long prefix conjugation in the first, the short one in the second and third persons—the so-called indirect volitive). This expresses consecutive or final action (e.g., Gn. 27:4, håḇi:'å li wǝ'oḵe:lå, “bring me in order that I eat”; Dt. 32:7, šǝ'al 'aḇiḵå wǝyaggedḵå, “ask your father that he may tell you”).

Circumstantial clauses resemble main clauses even more. Usually, they follow the main clause, connected by or asyndetically (e.g., Gn. 18:1, wayyerå 'ælåw…wǝhu yošeḇ, “and He appeared unto him … while he was sitting”). When circumstantial clauses precede the main clause, it is the main clause that is connected by , and it is only the internal structure of the circumstantial clause (the subject preceding the verb, nominal clause, etc.) that differentiates it from the main clause; Genesis 38:25–26, reflects a quite intricate sentence structure: hi muṣeṯ wǝhi šålḥåwatto:mærwayakker, “and while she was being brought forth [preceding circumstantial clause], and she had already sent [background clause, the subject preceding the finite verb] … and had said [continuation of the background clause, yet theoretically it could denote the main action as well], … and he acknowledged [main action]”). The number of subordinating conjunctions is relatively small. The main relative pronoun is 'ašær. In Hebrew, relative clauses differ in their structure from, for example, English ones, being much more independent. In their full form, pronominal suffixes, governed by a preposition/verb, refer back to the antecedent. In English it is the relative pronoun that is governed by the preposition/verb (e.g., 'ašærbo=“in which”). 'ašær is also used to introduce substantive clauses. The main conjunction for introducing substantive clauses is ki, which has a variety of functions, including emphasizing particle and causal and temporal conjunction. Important particles are the conditional 'im, hypothetical lu, and presentative hinne, “behold”; the latter is often followed by a participle marking the future, whereas wǝhinne, literally “and behold,” is often dependent on verbs of “seeing.” It marks what has been seen, as in Genesis 30:6: wayyar(') 'oṯåm wǝhinnåm zoap̄im, “and he saw that they were vexed.”


The vocabulary of the Bible (and even moreso of Hebrew inscriptions), is restricted; its language is also somewhat one-sided: it deals with religion, morals, and emotion, being the tongue of prophets, legislators, and poets. Small wonder that many semantic fields are poorly represented and that the number of hapax legomena is relatively high. Biblical vocabulary is especially influenced by Aramaic, whose impact increases in the book's later portions. Aramaic has also served as the intermediary for several Akkadian loanwords.

Second Temple Period.

The destruction of the First Temple (587 BCE and the Babylonian Exile marked a new stage in the development of Hebrew culture and language. The returning exiles brought with them the Aramaic language and culture. The books of the Bible had been written during the First Temple period by people whose spoken language, for all their differences, did not differ from the literary tongue more than any other spoken language differs from its literary layer. With the return of the exiles a very much different dialect prevailed in everyday speech that was more and more influenced by Aramaic. This dialect had existed in the First Temple period (reflecting an only marginal Aramaic influence), but people speaking it were, as a rule, cut off from the literary tongue. In the Second Temple period, however, people speaking this dialect were among the authors of literary works written in Late Biblical Hebrew, which, accordingly, reflected genuine diglossia. It was only later that the rabbis adopted this spoken dialect as a literary language and utilized so-called Rabbinic Hebrew, which became the second classical Hebrew tongue, after Biblical Hebrew.

The biblical books composed during this period—such as Chronicles, Ezra and Nehemiah, Esther, Ecclesiastes, and Daniel—are, for all their differences, written in Late Biblical Hebrew. Besides utilizing matres lectionis more frequently, the language prefers, inter alia, 'ani, “I,” to 'ånoḵi and tends to mark the direct object by , presumably through the impact of Aramaic. The latter might also have caused the reduction of the use of converted forms, and the increased use of noun patterns like malḵūṯ, “kingdom,” instead of the classical mamlāḵā.

The apocryphal books are also written in Late Biblical Hebrew. Fragments of them have been discovered in Hebrew in Jubilees, the Testaments of the Twelve Patriarchs, and Ben Sira. The Hebrew parts of the latter, a book belonging to the Wisdom literature and composed in about 180 BCE, were first discovered in the Cairo Genizah (the storeroom of the Ben-Ezra Synagogue in Cairo in which worn Hebrew books and documents were kept) and later in the excavations on Masada. The Dead Sea Scrolls, the manuscripts found at Qumran, near the northern edge of the Dead Sea), and in its vicinity, were, except for biblical texts, composed by Jewish sectarians. Their historical importance lies in the light they shed on the two centuries following the Hasmonean uprising, a period decisive for the history of monotheistic religions. [See Dead Sea Scrolls; Qumran; Masada; Hasmoneans.] These scrolls are written in another variety of Late Biblical Hebrew, dubbed Qumran Hebrew. It utilizes, inter alia, pronoun doublets (forms with and without final heh), bǝlō(')/lǝlō(') with finite verbs, and predicatively used infinitives. Yet, although the language of the scroll Miqṣat ma῾aśe ha-Torah from Qumran Cave 4 grammatically resembles Biblical Hebrew more than Rabbinic Hebrew, it is closer to the latter in its use of še, “that,” and in its vocabulary, perhaps because of its halakhic character. Nothing certain can be said so far about the Copper Scroll, which reflects affinities with Rabbinic Hebrew as well. [See Copper Scroll.]

Rabbinic Hebrew is to be divided into two main periods: tannaitic and Amoraitic. In the Tannaitic era (in the first two centuries CE.), Hebrew was still a spoken language, though rivaled by Aramaic and even by Greek. The tannaim composed all their works in Rabbinic Hebrew: the large collections dealing with law (halakha)—the Mishnah, the To-sefta, and the Baraitot (halakhic traditions not accepted into the Mishnah and Tosefta)—and the early Midrashim, which comprise both halakhic and homiletic (haggadic) exegesis. No early manuscripts written in Rabbinic Hebrew survive. The earliest specimen in a somewhat different species of Rabbinic Hebrew are the Bar Kokhba letters, some of which were written in Aramaic and even in Greek (c. 135 CE), found mainly in Naḥal Ḥever and Wadi Murabba῾at in the Judean Desert. [See Bar Kokhba Revolt; Judean Desert Caves.]

Later manuscripts were decisively influenced by Biblical Hebrew, so that the reconstruction of genuine Rabbinic Hebrew is somewhat precarious. The latter clearly differs from the former not only in vocabulary (partly the result of the Aramaic influence), but in grammatical structure as well. Converted verbal forms disappeared, as did the infinitive absolute. The infinitive construct coalesced with preceding it and in the basic verbal theme was restructured according to the prefix tense. In final position, the opposition of m:n was neutralized. The past is marked by the suffix tense, while the present and, as a rule, the future, are marked by the participle. The prefix tense is, to a great extent, used as a sort of subjunctive. The lengthened, and as a rule also shortened, forms of the prefix tense are no longer extant. The attributive use of the demonstrative pronoun is archaic when compared with Biblical Hebrew because it is not pleonastically determined by the definite article (Rabbinic Hebrew's bayiṯ ze, “this house,” as against Biblical Hebrew's hab-bayiṭ haz-ze). The number of subordinate clauses is much higher than in Biblical Hebrew, and most are built with the conjunction še.

It was in the wake of the collapse of the Bar Kokhba Revolt in 135 CE that Hebrew ceased to be a living language and was used for cultural (religious) purposes only—though it often flourished as a literary tongue. It was not spoken again until the end of the nineteenth century in Israel (Palestine). Because Biblical and Rabbinic Hebrew, despite the differences between them, continued to serve as cultural unifiers—Bible and Mishnah being the two main pillars of Judaism—once Hebrew ceased being spoken, various mixtures of Biblical and Rabbinic Hebrew were utilized in every historical period. Each of these dialects was influenced by the spoken non-Hebrew vernacular wherever Jews lived. In the Hebrew of later periods, significant archaizing tendencies can also be detected. Amoraitic Rabbinic Hebrew (c. 200–500 CE), attested in the later Midrashim, as well as in the primary source of Jewish law, the Palestinian and Babylonian Talmuds (compiled in the first half of the fourth and the second half of the fifth centuries, respectively), already reflects the mixture of the two, although Rabbinic Hebrew is its main component. It is decisively influenced by the vernacular—Aramaic—in which a great part of the Talmuds and the late Midrashim are written.

Similar mixtures of Biblical and Rabbinic Hebrew and the vernacular are reflected in the medieval varieties of literary Hebrew. Thus, Spanish Hebrew poetry, in the main, imitates Biblical Hebrew, yet it is not entirely devoid of elements belonging to Rabbinic Hebrew; moreover it is also influenced by the vernacular—Arabic.

Modern Hebrew.

The revival of Hebrew as a living language in Modern (Israeli) Hebrew is the only case in the history of languages in which a dead language (i.e., one used only for literary purposes) again became a spoken tongue. The coincidence of three factors at the beginning of the twentieth century in Palestine (later the State of Israel) made the revival possible. Many of the Jews in Palestine had emerged from traditional Jewish society, which taught its members the Hebrew language as a part of Jewish culture, thus imparting to them a working knowledge of the language. Because these Jews used various vernaculars, Hebrew was their only common language. Many of the Jews in Palestine additionally were influenced by the zeitgeist of nationalism, which pleaded for the use of the national language in every domain. This provided the members of the Zionist movement with the motivation to revive Hebrew. It was through the Hebrew Language Committee (later the Academy of the Hebrew Language) that missing words, especially technical terms, were created. The morphology of Modern Hebrew is in the main identical with that of Biblical Hebrew. Yet, from doublets forms were chosen that occur in Rabbinic Hebrew as well. Its sentence formation is much closer to Rabbinic Hebrew, but it is also influenced by Standard Average European, in which the number of subordinate clauses is also much higher than in Biblical Hebrew. Standard Average European has also considerably influenced Modern Hebrew phraseology.

[See also Biblical Literature, article on Hebrew Scriptures.]


  • Bauer, Hans. “Die Tempora im Semitischen.” Beiträge zur Assyrologie 8.1 (1912): 1ff.
  • Bauer, Hans, and Pontus Leander. Historische Grammatik der hebräischen Sprache des Alten Testaments (1922). Hildesheim, 1962.
    Linguistically sound approach, but sometimes impaired by an excessive desire for innovation; includes phonetics and morphology only
  • Ben-Ḥayyim, Zeev. The Literary and Oral Tradition of Hebrew and Aramaic amongst the Samaritans, vol. 5, Grammar of the Pentateuch (in Hebrew). Jerusalem, 1977.
    Authoritative grammar of Hebrew according to the Samaritan tradition, including phonetics and morphology
  • Bergsträsser, Gotthelf. Hebräische Grammatik. 2 vols. Leipzig, 1918–1929.
    Although a torso, containing only phonetics and the morphology of the verb, and somewhat dated, it is still the best Hebrew grammar.
  • Bergsträsser, Gotthelf. Introduction to the Semitic Languages (1928). Translated by Peter T. Daniels. Winona Lake, Ind., 1983.
    Masterly, though somewhat antiquated, introduction to Semitic linguistics that also contains a superb description of Biblical Hebrew.
  • Blau, Joshua. A Grammar of Biblical Hebrew. 2d ed. Wiesbaden, 1993.
    Short grammar expatiating on problems referred to in this entry.
  • Brown, Francis, et al. A Hebrew and English Lexikon of the Old Testament. Corr. ed. Oxford, 1952.
    Superb, though somewhat obsolete, dictionary of the Bible that can only be compared with Buhl's dictionary (see below).
  • Buhl, Frants. W. Gesenius' hebräisches und aramäisches Handwörterbuch über das Alte Testament. 17th ed. Leipzig, 1921.
    Superb, though somewhat outdated, dictionary of the Bible, comparable to Brown's (above).
  • Cohen, Harold R. Biblical Hapax Legomena in the Light of Akkadian and Ugaritic. Mīssoula, 1978.
    The most reliable treatise on hapax legomena.
  • Davidson, A. B. Hebrew Syntax. 3d ed. Edinburgh, 1924.
    Clear, sound work.
  • Driver, Samuel R. An Introduction to the Literature of the Old Testament (1891). New York, 1963.
    Still one of the best introductions to the Bible, with superb linguistic analysis.
  • Driver, Samuel R. A Treatise on the Use of the Tenses in Hebrew and Some Other Syntactical Questions. 3d ed. Oxford, 1892.
    Even if one disagrees with the author's conception of the Hebrew tenses, it is possible to profit greatly from this masterful collection of material and clear exposition.
  • Elliger, Karl, and Wilhelm Rudolph. Biblia Hebraica Stuttgartensia. Stuttgart, 1968–.
    The best-known Western edition of the Bible; its apparatus criticus should be used cautiously.
  • Joöon, Paul, and Takamitsu Muraoka. A Grammar of Biblical Hebrew. Rome, 1991.
    The most up-to-date, comprehensive grammar.
  • Kautzsch, Emil. Gesenius' Hebrew Grammar. Translated by Arthur E. Cowley. 2d ed. Oxford, 1910.
    Though obsolete and often linguistically unsound, the collection of material is of great importance.
  • Köhler, Ludwig. Hebräisches und Aramäisches Lexikon zum Alten Testament. 3d ed., revised by Walter Baumgartner et al. Leiden, 1967–.
    The most up-to-date dictionary of the Bible, although not up to the standard of Brown's and Buhl's dictionaries (see above).
  • Kutscher, Eduard Y. A History of the Hebrew Language. Jerusalem, 1982.
    Unfinished original account by one of the foremost Hebrew linguists.
  • Sáenz-Badillos, Angel. A History of the Hebrew Language. Cambridge, 1993.
    Well-balanced history of the Hebrew language.
  • Waltke, Bruce K., and Michael O'Connor. An Introduction to Biblical Hebrew Syntax. Winona Lake, Ind., 1990.
    Up-to-date treatment, with a good bibliography; the conclusions should be used cautiously.

Joshua Blau