HEBREW LANGUAGE


This entry is arranged according to the following scheme:

A detailed table of contents precedes each section.

PRE-BIBLICAL

Pre-biblical Hebrew, for the purposes of this article, refers to the Hebrew language as reflected in written documents up to and including the 12th century B.C.E. (Taanach, cf. below). This study is limited to an analysis of the written evidence which reflects the Hebrew spoken in Palestine during the pre-biblical period. (No attempt will be made to reconstruct Proto-Hebrew (‡) forms based on the masoretic Hebrew found in the Bible). Evidence of this type exists for the period from the mid-20th to the 13th centuries B.C.E.

Nature of the Evidence

There is no corpus of texts written in pre-biblical Hebrew, but only toponyms and single words transcribed into syllabaries which were not able to render accurately the consonants and vowel patterns of Hebrew. This material is written in the Egyptian and Akkadian syllabaries and care must be exercised in reconstructing the language upon which these transcripts are based.

The Akkadian syllabary (as a result of Sumerian influence) was unable to represent gutturals except for . Therefore, these transcriptions cannot help to determine which of the Proto-Semitic gutturals were still in use in pre-biblical Hebrew. However, whenever a guttural was pronounced it was usually rendered by an , e.g.,

נְחֹשֶׁת nu-ḫu-uš-tum (EA 69:28)

עָפָר ḫa-pa-ru (EA 141:4; also a-pa-ru, EA 143:11)

זְרוֹעַ zu-ru-uḫ (EA 287:27)

צֹהַר zu-uḫ-ru-ma (EA 232:11)

The Egyptian material, while more faithfully representative of the consonantal system, did not have a clearly defined vowel system. As far as the vocalization of the Egyptian material is concerned, this study bases itself on W.F. Albright's Vocalization of the Egyptian Syllabic Orthography (VESO).

In discussing the verbal scheme of pre-biblical Hebrew, it must be carefully determined if a particular form is pure Canaanite, or whether it is in fact simply poor Akkadian, or the result of contamination – a combination of Akkadian and Hebrew elements (cf. the discussion on yaqattal forms, Morphology no. 5 below). When assessing place names, it is necessary to remember that they are conservative by nature, and do not always undergo the same linguistic changes as other words (e.g., Ak-ka Phonology no. 5). It is important, too, to remember that place names are often lexically difficult and may at times be non-Semitic in origin (cf. Hazor Phonology no. 5). Naturally, such words are expected to behave differently than the norm of pre-biblical Hebrew. Finally, it must be remembered that Canaanite itself was probably under the strong influence of Amorite, and that some forms found in the transcriptions may be directly due to this influence.

Keeping these factors in mind, and remembering that there is no corpus of pre-biblical Hebrew as such, but words or groups of words written in non-Hebrew syllabaries, it is surprising how much can actually be said about the phonology and morphology of the earliest stage of Hebrew.

The Sources

(1) The Egyptian material consists of lists of Canaanite personal and place names. In 1909, Burchardt published all the words and Hebrew parallels known at that time. Much more important are the Execration texts, published first by Sethe, and supplemented by Posener. These are texts inscribed on vases and contain the names of potential rivals; it was assumed that with the smashing of the vase, the opponent would also be destroyed. These texts date from the mid-20th to the late 19th centuries B.C.E. The Egyptian material is completed by an 18th-century list of Egyptian slaves (The Hayes List) which contains more than 30 North-West Semitic names, and was published by Albright. In all, there are 150 names from the period between 1900 and 1750 B.C.E.

(2) At Taanach (5 mi. (8 km.) south of Megiddo), some important cuneiform material was unearthed by the archaeologist E. Sellin in 1903–04. It contains six letters written in cuneiform Akkadian, which date from the 16th–15th-century B.C.E. period and were edited by the Assyriologist F. Hrozny in 1904. The letters appear to have been written by Canaanite scribes (especially the first two), and in several instances their native speech is clearly reflected in the strange Akkadian forms found in the letters. This material was supplemented by another letter found at Taanach and published originally by D.R. Hillers in 1964; it is written in alphabetic cuneiform and dates from the 12th century B.C.E. (cf. F.M. Cross's republication of this letter).

(3) It has long been recognized that many of the Ur III and Old Babylonian names found in Akkadian sources do not reflect standard Akkadian, but a distinct dialect or language. The traditional division of Semitic languages sets Akkadian off as East Semitic, and it was naturally agreed that these names, if they are not Akkadian, must be West Semitic. In 1926, Th. Bauer collected and analyzed about 700 names of this type in his work Die Ostkanaanaeer and argued (following Landsberger) that these names were Canaanite (but originated east of the Tigris, hence were East Canaanite). Since many of these names were prefaced by the Sumerian ideogram MAR. TU, which is equivalent to the Akkadian amurru ("west"), the people who bore these names came to be known as Amorites. Other scholars (Albright) have suggested that the language was a dialect (eastern) of Canaanite, while some (Goetze) feel that Amorite and Ugaritic make up a separate division within North-West Semitic. Since Bauer's publication, the number of known Amorite names has virtually tripled; a more up-to-date study can be found in Huffmon's Amorite Personal Names in the Mari Texts (with a complete bibliography of the problems of Amorite). It is not clear whether the designations MAR. TU (Sumerian) and amurru (Akkadian) were originally independent terms or linguistic equivalents. In the older (Ur III) period, persons described as MAR. TU have names which are often not Semitic, while in the later (Old Babylonian) period, many persons with West Semitic names are not designated in this way. This fact has prompted scholars to distinguish two linguistic strains: one called Amorite (Ur III) and the other East Canaanite. It seems clear today that the relationship between these two groups is much closer than originally thought and both may be referred to as Amorite (cf. Gelb, in JAOS, 88 (1968), 39–47).

The similarity between many Amorite names and names found in the Egyptian Execration texts, points to the importance of this material for the study of pre-biblical Hebrew. There undoubtedly were many points of similarity between the Amorite West Semitic dialect and the language of Canaan; Amorite, however, is not Canaanite as seen clearly from the fact that ā does not go to ō (the Amorite s = the Canaanite š is also significant).

(4) Possibly the most important source which has direct bearing on the language spoken in Canaan in pre-biblical times are the *Tell el-Amarna letters (EA). Discovered in 1887 at Tell el-Amarna, in Egypt, these letters contain the correspondence between the ruling Egyptian court and their vassal princes in Canaan. The letters are ostensibly written in Akkadian, which was at that time (14th century B.C.E.), the lingua franca of the Near East. However, the Canaanite scribes were not fully conversant with the Babylonian language they attempted to write, and constantly substitute Canaanite forms and idioms for the Babylonian, producing a real "Mischsprache" which, when analyzed, yields much information about the scribes' native tongue. There is, in fact, at least one Canaanite proverb written entirely in the language of the scribe: ki-i na-am-lu tu-um-ḫa-zu la-a ta-ka-bi-lu u ta-ansi-ku qa-ti amelim ša yi-ma-ḫa-aṣ-ṣi: (meaning, "If ants are smitten, they do not accept [the smiting quietly], but they bite the hand of the man who smites them") (EA 252:16–19; cf. Albright, in BASOR, 89 (1943), 29–32).

More important than the native influence on the Akkadian in these letters are the Canaanite glosses which the scribe wrote in the margins of the text (in the Akkadian syllabary) as the equivalents to Akkadian words in the body of the letter. For example, in a letter from Jerusalem (EA 287:27), the Akkadian qat has the gloss zu-ru-uh is equivalent to the Hebrew zәroaʿ (זרוע, "hand"). In the Akkadian syllabary ʿ = ; and ō, which was non-existent, was transcribed as a u; therefore, zu-ru-uḫ is the exact equivalent to the Hebrew זרוע.

The reasons for writing these glosses are unclear. The scribe may have written the Canaanite gloss for an Akkadian word he was not sure of, and filled in the proper form later. This explanation is the most likely since the gloss is always the exact equivalent of the Akkadian word in the text, and the scribe surely did not expect the gloss to clarify the text for the Egyptian reader. Still, whatever the reason, these glosses remain the most direct evidence of pre-biblical Hebrew.

The letters were given an excellent edition by Knudtzon (with some help by Weber and Ebeling) in 1915. Further texts were published by Dossin, Gordon, Schroeder, and Thureau-Dangin (for complete references cf. R. Borger, Handbuch der Keilschriftliteratur, 1 (1967), 238). This was supplemented by fine grammatical studies by Boehl, Ebeling, Dhorme, Albright, and Moran (cf. bibliography).

The material is presented under two headings: Phonology and Morphology. Each grammatical point is given separate treatment with the appropriate title preceding the evidence. When necessary, any differences between the earlier (Egyptian) and the later (cuneiform) material are noted and explained.

Phonology

(1) AND These two Proto-Semitic consonants are not distinguished in the Hebrew script. The Phoenician-Canaanite alphabet contains only one grapheme for these two consonants indicating that the language for which this alphabet was developed did not differentiate between these two sounds. The Tiberian vocalization shows that the reading tradition also did not distinguish between the two sounds, as a diacritic was not used to differentiate between them (as opposed to š and ś). Greek material indicates some differences in transcribing the ח, but this seems to be rooted in secondary and dialectical developments. It is, therefore, necessary to look elsewhere to determine whether Hebrew ever differentiated between the and . The material written in cuneiform (Taanach, Amorite, Amarna) is of no help here since the Akkadian syllabary only recognizes the consonant ; thus the distinction, if it did exist, could not be represented in that syllabary. The Egyptian material, however, is of crucial importance.

The early Egyptian material clearly distinguishes between and :

ḥa-ar-pu in Canaanite ‡ḥarbu, Hebrew ḥereb (VESO XII, A, 4); while

ḫu-ru in Canaanite ‡ ḫurru, Hebrew ḥor (VESO XIII, A, 5).

In the later Shoshenq list (c. 950), the etymological seems to have merged with , as: bt ḥ(w)rn (Beth-Horon) (which etymologically is probably ). This indicates that the Egyptian (Amarna) documents reflect the state where the assimilation of to was becoming finalized, and that this later state is reflected in the Shoshenq list.

It is difficult to accept the position (Goetze) that where these consonants are differentiated, it is as a result of Amorite (!) influence. Amorite was written in the cuneiform syllabary which could not differentiate between and , so there are no objective grounds for assuming that Amorite made this distinction. This position can only be justified by the assumption that Ugaritic and Amorite form a subgroup within North-West Semitic, and that the Ugaritic differentiation was maintained in Amorite. But there is no real evidence that this distinction was maintained in Amorite, and it is more reasonable to assume that cuneiform writing limited the Amorite pronunciation of gutturals just as it did the Akkadian. While there was clearly an Amorite influence in Canaan, there is no reason to assume that some early stage of Hebrew could not have differentiated between and as in the early Egyptian material.

(2) INDEPENDENCE OF THE Ś PHONEME IN PRE-BIBLICAL HEBREW The fact that only two graphemes are used in biblical Hebrew to distinguish the three sounds s, ś, š, (ś and š are distinguished by means of a diacritic) reflects the situation in the language from which the alphabet was borrowed, and not directly on Hebrew. The fact that Greek and Latin transcriptions were unable, at a much later date, to distinguish between the three sounds is not relevant, since there is only one Greek-Latin grapheme s (σ) which parallels the early Hebrew sibilants.

In biblical Hebrew the three phonemes s, ś, and š are kept distinct. This is also the case in Epigraphic South Arabic, while in the other Semitic languages these three phonemes have coalesced into two. This indicates that the situation reflected in biblical Hebrew is primary, and not the result of a late innovation.

In the pre-biblical cuneiform material, this distinction is difficult to recognize (where only s and š are distinguished), while the Egyptian material seems clearly to indicate that the distinction between these three phonemes was carefully maintained. Here š and were transcribed by the sign for s, while š was represented by š, so that etymological ś and š were kept distinct:

Ya-si-r-ʾi-ra, in Hebrew ‡Yasirʾel (ישראל, "Israel") (VESO III B3);

sa-ʿa-ra-ta, in Canaanite ‡śaʿar (a)t(a), in Hebrew saʿara (שערה, "hair") (VESO V A 13); while,

ša-ʿa-ra, in Hebrew šaʿar (שער, "gate") (VESO XV A 4).

(3) AND š The Hebrew-Phoenician alphabet has only one grapheme for these two Proto-Semitic consonants, and they both seem to have coalesced into š.

The early Egyptian material seems to indicate that the distinction between these two phonemes was still maintained. In these transcriptions both and ś are written with an s:

ʾa-ti-ra, in Canaanite ʾasira (a), in Amarna asiru, in Hebrew ʾasir (אסיר);

ṯu-pi-ir, in Canaanite soper, in Hebrew sofer (סופר), but ʿstrt (ʿaṯtartu) = Astarte (Burchardt 285);

ḥa-da-sa-t (hadaṯatu), in Canaanite hadaš (a)t(a), in Hebrew ḥadaša (חֲדָשָׁה, "new") (VESO XII, A, 6).

But š is transcribed as š in the word qa-di-š, in Hebrew qadoš (קדש) (VESO XC9).

However the same lists mention ša-ʿa-ra, in Canaanite ‡ša-ar, in Hebrew šaʿar (שער, "gate") (VESO V, A, 14) (from ṯaʿaru), indicating that the was unstable even here and that it tended to merge into š. In the Shoshenq list, the š and are undifferentiated: šbrt = šibbolet (from tibbolet?). In short, the was already on its way to coalescing with the š in the period of the early material, and by the time of the Shoshenq list and biblical Hebrew, it had disappeared as a separate phoneme.

The Amarna evidence is difficult to interpret. On the one hand, the difference in spelling between La-ki-si (EA 288:43) and Ša-ak-mi (EA 289:23) in the Jerusalem letters indicates that in that area š and were kept etymologically distinct (š in Ša-ak-mi must be from , cf. Ugaritic ṯkm, as well as the fact that it is written with an s in the Egyptian transcriptions (VESO XIV A 15); and š in masoretic Hebrew). However, it has been noted (Goetze) that the spelling La-ki-ši also appears in the Jerusalem letters (EA 289:13) and that the cuneiform signs for sa, si, su, serve as scribal variants for ša, ši, šu.

(4) ʿayin and gyin These two phonemes, which have coalesced into ʿayin in biblical Hebrew, are still distinguished in the Egyptian material.

ʿn-qn-ʿa-m(a), in Canaanite En-qne-ʿ-am(ma), in Hebrew Yoqněʿam (יקנעם) (VESO V, A, 6), while

ʿa-da-ta, in Canaanite ‡Gazzat(a), in Amarna, Azzati, Hazati, in Hebrew Azzah (עזה, Gr. Γαζα) (VESO XVI, A, 11).

However, we also find ša-ʿa-ru, in Hebrew שער ("gate") < ṯaεru (VESO V, A, 14), which would indicate that g > ʿ, as it has in biblical Hebrew. Here too, it must be pointed out that while the transcriptions prove that the distinction between the two phonemes still existed in early Hebrew, the distinction was fast disappearing. The cuneiform material can be of no help as the syllabary does not (generally) distinguish between these consonants.

In a published letter from Taanach (BASOR, 173 (1969), p. 45–50), dated the late 13th or early 12th century B.C.E., the name pʿm (puʿm) appears. If this name is related to the Ugarit pgm, this would indicate that g < ʿ. (This letter is written in an alphabetic orthography (related to Ugaritic?), and it may be assumed that if the g = ʿ distinction still existed in Canaanite, it would have been represented in this way).

(5) LONG a (ā) BECOMES LONG o (ō) All the relevant material indicates that this shift, which is considered unique in Canaanite (of the classical Semitic languages), took place as early as the 15th century B.C.E. An instance from Amorite is especially interesting. The name of a northern city appears as Ḫaṣura (Hazor). Since the ā > ō shift did not take place in Amorite at all, this seems to indicate that the situation in Canaan in the 15th century B.C.E. was already post shift. (However, care must be taken in this case since the name is etymologically unclear and its vocalization may reflect a non-Semitic pronunciation.) All other relevant material indicates that when the Amarna period started the shift had already taken place in Canaanite.

(a) Egyptian Material. Bi-ʾa-ru-ta, in Canaanite beʾrot, in Hebrew באר (VESO X, C, 4)

(b) Amarna Canaanite (Many Examples). a-nu-ki (EA 287:66, 69), in Hebrew אנכי; ṣu-un-nu (EA 263:12), in Hebrew צאן; ru-šu-nu (EA 264:18), in Hebrew ראשנו. The place name Ak-ka, in Hebrew עכו reflects a conservative pronunciation.

(c) Taanach. Interestingly enough the only indication for this shift in Taanach is the name Gu-li which may be derived from the Canaanite Goʾeli, Hebrew גואל (I 3) (but cf. below). However, other place names, such as Ra-ḫa-bi (IV 22), Hebrew רחוב, and Ma-gi-id-da (V 15), Hebrew מגדו, seem to indicate that this shift had not yet taken place. (Another possible reading for the name in IV 22 is Elu-ra-pi-i, Hebrew רפא.)

Unfortunately, all the Taanach evidence is in the form of personal and place names, and it may be that although the shift had taken place generally, these names had preserved an older pronunciation. It is also possible that the shift had not yet taken place in the pre-Amarna period, or, at any event, not in the south of Canaan. The only evidence to the contrary is the name Gu-li which may be part of the Canaanite Gu-li-Adad (Albright), or the Hurrian Guli-Tešub (Maisler (Mazar)), in which case it is not relevant to the problem. The sum of the evidence seems to indicate that in Taanach, in the period preceding Amarna (16th–15th century) the shift ā > ō had not yet taken place.

(6) THE PROTO-SEMITIC n IS ASSIMILATED TO THE FOLLOWING CONSONANT In biblical Hebrew, the Proto Semitic n is assimilated to the following consonant. The pre-biblical material shows that this process was in a state of flux and that in the early period the n was not as consistently assimilated as later on.

The Egyptian Bi-in-ti-ʿ-n-t is in Canaanite ‡Bint (i)-ʿanat ("daughter of Anat"; Canaanite bint = Hebrew בת). In other words the n of בת remained unassimilated (VESO VI B 12).

In Amarna there is both gitti < ginti as well as ginti (Hebrew gat), indicating that the tendency to assimilate the n did exist, but had not yet reached the proportions of biblical Hebrew where it is the norm in these circumstances. La-bitu < labintu (EA 296:17) is also found in Amarna.

(7) aʾ > (ā) > ō The Amarna material indicates that this change had already taken place in Hebrew:

(a) ru-šu-nu (EA 264:18), roš (ראש, "head") reflects the following development: rōš < rāš < ‡raʾš.

(b) zu-u-nu (EA 263:12); ẓon (צאן, "sheep") here too we assume zōn < zān < zaʾn.

The Egyptian material generally reflects the later stage, as in the name ršqdš ( = ראש) (VESO X, C, 9), but ru-ʾu-š (a) is also found (VESO X, C, 7). Ru-ʾu-š(a)-qdš, in Canaanite ‡Rošqids, is explained by Albright as being a Canaanite back formation of roʾ(o)š from rōš, on the analogy of the plural raʾšim (ראשים) which was probably dialectic (VESO III, E, 6). Another possible interpretation is that this name preserved the earlier pronunciation (cf. the name Ak-ka (עכו) in Amarna Hebrew (Phonology, no. 5).

(8) MIMMATION Mimmation, known in early Akkadian, is almost unknown in biblical Hebrew, except for a few fixed forms like חנם ,יומם ,אמנם. The earlier material preserves several interesting examples. In the Execration texts, the city of Jerusalem is transcribed as ვwsვmm (= Urusalimum). Similarly, the Taanach letters indicate a high rate of mimmation, especially in the Canaanite letters. (In Amarna there is a marked decrease in the use of mimmation, an indirect indication that the Taanach letters precede the Amarna letters).

Morphology

(1) CASE ENDINGS The Semitic languages had originally three basic cases: nominative, accusative, and genitive, which were differentiated by suffixes (in singular -u,-a,-i). Especially prominent in the early history of the language, they generally fell into disuse. In biblical Hebrew, the nominal case endings have generally been lost, except for certain compound names like מתושלח, and possibly poetic forms like בנוֹ בעור, and חיתוֹ ארץ. The heh localis, long thought to be a preserved accusative in biblical Hebrew, has to be reevaluated in the light of the Ugaritic evidence. In Ugaritic, which does not generally employ the mater lectionis, there is the word šmmh which would indicate that the h has consonantal force (cf. Speiser, in: Israel Exploration Journal, 4 (1954), 108–115).

Both the Egyptian material and the cuneiform material from Amarna indicate that case endings were still in use at that time. In the earlier Egyptian lists, place names usually end with -u, but the tendency was for -a to replace -u in later material, e.g., Ayaluna, Ašqaluna, Ṣiduna, and Ḫaṣura. This indicates that before the case endings were completely dropped, there was a period where these final short vowels were confused and the cases were no longer grammatically distinguished.

Boehl has shown (see also Dhorme, in: RB, 23 (1914), 347–8 = Recueil E. Dhorme (1951), 460–1) that the case endings in Canaanite are by and large not confused and the distinction between the different cases is maintained. Examples of the use of the nominative -u:

a-pa-ru (עפר) EA 141:4

ru-šu-nu (ראשנו) EA 264:18

The genitive -i:

sa-aḫ-ri (שער) EA 244:16

a-na-yi (אֳנִי) EA 245:28

The accusative -a:

ḫa-an-pa (חנף) EA 288:7

mu-ur-ra (מֹר) EA 269:16

These case endings were elided in Hebrew after the Amarna period and are very rare in biblical Hebrew.

(2) THE FIRST PERSON SINGULAR INDEPENDENT PRONOUN: ʾanoki The ʾanoki form is found in Amarna Hebrew (a-nu-ki EA 287:66, 69). In late biblical Hebrew, this form becomes rare and is almost always replaced by ʾani.

(3) THE DUAL ENDING: -ay (m)a In Amarna, the dual ending is known from the word hi-na-ia ("my (two) eyes" (nom.)) (EA 144:17). A similar form is known from Taanach iṣma-ga-re-ma (II, 8) ("two Chariot wheels"), with the Canaanite dual endings.

(4) FIRST PERSON POSSESSIVE SUFFIX: -nu (OUR) Timi-tu-na-nu ((and you) "killed us"), in Amarna (EA 238:33).

THE VERB

(1) CAUSATIVE PREFIX ha-. The Hebrew causative prefix ha- appears in Amarna as ḫi- (attenuation). The example is from EA 256:7, ḫi-iḫ-bi-e; it is clearly a Hebrew form which is impossible in Akkadian. The scribe used the Hebrew החביא for the common Akkadian verb of the same meaning, puzzuru.

(2) THE ENDING OF THE FIRST PERSON SINGULAR OF THE QATALA FORM: -ti, qatalti. In Amarna, qatalti forms appear, e.g., ba-ni-ti "I built" (EA 292:29).

TENSES

The Hebrew verbal scheme, which consists (primarily, cf. below) of two tenses, differs radically from the Proto-Semitic system. The two tenses: a prefixed one indicating incompleted action (imperfect), and the suffixed one indicating completed action (perfect), are secondary. It is possible to give approximate dates for the introduction of these verbal forms from the cuneiform evidence for early Hebrew.

(3) QATALA PRETERITE FORM The qatal(a) form, which in Akkadian is the basis of the stative form (a nominal not a verbal form in Akkadian), is found in Amarna serving also as the Hebrew preterite form. For example:

la-ma-ad ("he has learned") (EA 196:30).

ša-al ("he questioned") (EA 289:10).

(4) YAQTUL IMPERFECT FORMS The Amarna evidence points clearly to the fact that the yaqtul imperfect form had already been developed in Canaan. However, as Moran has pointed out, this function was not exclusive with this form. His work on the Amarna letters from Byblos led him to the conclusion that the imperfect indicative had two functions: (a) present future; (b) past iterative. The passage EA 104:17–36, which Moran quotes, is instructive. The verb lequ ("take") appears three times in the prefixed form tilquna, once as a present, once as a future, and once as a past.

miya mārū Abd-aširta ardi kalbi šar māt Kašši u šar māt Mitanni šunu u tilqûna māt šarri ana šāšuna panânu tilqûna ālāni ḫazānīka u qâlāta annû inanna dubbirū rābiṣaka u laqû ālānišu ana šāšunu anumma laqû al Ullaza šumma kiʾamma qâlāta adi tilqûnu āl Ṣumuru u tidûkūna rābiṣa u ṣāb tillati šaina Sumura.

Who are the sons of Abd-aširta the slave and dog? Are they king of the Kassites or the king of the Mittani that they take the royal land for themselves? Previously they used to take the cities of your governors, and you were negligent. Behold! now they have driven out your commissioner and have taken his cities for themselves. Indeed, they have taken Ullaza. If you are negligent this way they will take Simyra besides, and they will kill the commissioner and the auxiliary force which is in Simyra.

(5) YAQATTALU FORM OF THE VERB The yaqattal form of the verb, attested to in the Akkadian present and the Ethiopic indicative, is generally thought to be missing from North-West Semitic. There are, however, certain forms found both in Amorite and in Amarna Hebrew which seem to indicate that an independent yaqattal form may have existed in North-West Semitic.

In Amarna Hebrew, the forms which seem to be of the yaqattalu pattern are, e.g., tidabbibu (EA 138:49)("they speak") and i-paṭ-ṭar (EA 2:46) ("he loosens"). However, the situation is not so simple. Even if the gemination in these forms is accepted as being genuine, they can be explained as Akkadian present forms with the Amarna Canaanite i- prefix in place of the Akkadian a-. In other words, these would not be genuine Canaanite forms but blends of Akkadian stems with Canaanite prefixes which were produced when speakers of Canaanite tried to write Akkadian.

In Amorite there are also certain names which give the impression of being yaqattal forms: e.g., the names yabanni (Akkadian: ibanni) and yanabbi (Akkadian: inabbi). Huffmon feels that these names can be properly compared to Ugaritic (cf. piʿel of bāna in post-biblical Hebrew), and do not necessarily reflect North-West Semitic yaqattal forms. Personal names, such as I-ba-as-si-ir and Ya-ba-an-ni-AN, are in his opinion clearly piʿel forms. There is therefore, in his opinion, no unambiguous evidence which indicates that there were in fact yaqattal forms of the verb in early North-West Semitic.

However, it has been pointed out (Von Soden, in: Die Welt des Orients, 3 (1964), 180) that the situation in Amorite is not at all clear:

(a) It is noteworthy that all the yaqattal forms that have been found in Amorite are derived from three weak verbs (excepting ibassir which may not even be Canaanite);

(b) Some of the roots, like yanabbi ("he calls") and yabanni ("he builds, makes") are not found as piʿel forms in Akkadian. The obvious conclusion is that these are not piʿel forms at all but the present of a true yaqattal form. (Some of the other names which appear in Amorite: yabassi, yaḫaṭṭ/ṭṭi, yamaṭṭ/ṭṭi, and yasaṭṭi are lexically unclear, and cannot be used as evidence.)

The fact that this form is clearly found in a North-West Semitic dialect (Amorite) may indicate that the Amarna Canaanite material should be reassessed. It may be that those forms which appeared to be Canaanite-Akkadian contaminations are true Canaanite Hebrew forms.

(6) YAQATULA (SUBJUNCTIVE) FORM OF THE CANAANITE VERB In biblical Hebrew, the moods of the imperfect stem have a limited use (opposed, for example, to classical Arabic). The situation in pre-biblical Hebrew is more difficult to determine. Since Akkadian has a homophonous morpheme, known as the ventive (a modal suffix in Akkadian which indicates motion toward the speaker or focus of attention; e.g., illik ("he came"), illikam ("he came here")), a large number of relevant occurrences of the suffix -a (with ya/i/u/qtal forms) are readily explained, at least at first glance as examples of the ventive.

Moran has proposed that true yaqatula forms are identifiable by semantic means and that they are specifically used in two instances (Orientalia, 29 (1960), 1–19):

(a) to express a wish, request, or command – yi-sa LUGAL ("may the king come forth").

(b) in clauses of purpose or intended result – ib-lu-ṭa ("so that I may live").

The fact that the biblical Hebrew form, known as the cohortative, has the same functions as the Byblian Amarna (a group of Amarna letters from Byblos) yaqtula, indicates that the Hebrew cohortative is a continuation of this "subjunctive" (H. Bauer and P. Leander, Historische Grammatikder hebraeischen Sprache (1922), 273; P. Joüon, Grammaire de l'Hébreu Biblique (1923), 315 n. 1).

(7) TAQTALU(NA) THIRD PERSON MASCULINE PLURAL FORM There seems to be some evidence from Amarna which points to the fact that a taqtalu (na) third person masculine plural form existed in pre-biblical Hebrew, as in Ugaritic (cf. Boehl, p. 53 and Moran, in: Journal of Cuneiform Studies, 5 (1951), 33–35). Hints of this form are also found in biblical Hebrew: e.g., ותקרבו in Ezekiel (37:7) as well as ותקרבון and ותאמרו in Deuteronomy (5:21).

(8) INFINITIVE ABSOLUTE USED AS A FINITE VERB In biblical Hebrew there are a few clear cases of the use of the infinitive absolute form of the verb with finite force. ושבח אני (Eccles. 4:2) ונהפוך הוא (Esth. 9:1) are the clearest examples of this phenomenon, paralleled by the qtl/yqtl ʾnk construction found especially at Karatepe. The fact that this was a fairly regular construction in Early Hebrew (and North-West Semitic in general) is shown by the many examples in the Amarna letters: e.g., (from Amarna), u ma-ti-ma šu-ut ("and when he died truly") where the form matima is an infinitive absolute followed by the independent pronoun with the force of an independent verb (cf. Moran, in Journal of Cuneiform Studies, 4 (1950), 169–72).

[Chaim Brovender]

The Names

Biblical Hebrew is also called early Hebrew in contradistinction to living middle Hebrew, as reflected in the Mishnah, the older portions of the Talmud and early Midrash; to written middle Hebrew, as it was used after its extinction as a living language; and to modern Hebrew. It is known mainly from the Hebrew portions of the Bible (i.e., the whole Bible with the exceptions of its quite restricted *Aramaic parts, i.e., two words in Gen. 31:47; Jer. 10:11; Dan. 2:4–7:28; Ezra 4:8–6:18), which constitutes a rather limited corpus. The name "Hebrew Language" itself does not occur in the Bible; instead, it is called the "language of Canaan" (Isa. 19:18) and "Judean" (II Kings 18:26, 28; Isa. 36:11, 13; Neh. 13:24, in the last passage already in accordance with the late, post-exilic usage, which extended the term "Judean" to the nation). Not until about 130 B.C.E. (in the prologue to Ben Sira) does εβραïστι – occur to denote old Hebrew. Josephus and the New Testament, however, use this term both of Hebrew and Aramaic (in contradistinction to Greek).

Ancient Evidence (Inscriptions and Transcriptions)

Only slight additional material can be adduced from inscriptions and transcriptions, which is not only due to their limited extent. The most important old Hebrew inscriptions are the calendar of *Gezer (c. 10th century B.C.E.), ostraca of *Samaria (from the eighth (?) century B.C.E.), the inscription of *Siloam (c. 700 B.C.E.), the *Lachish and Tel *Arad letters (sixth century B.C.E.). Their linguistic evaluation is impeded by their consonantal script, even the vowel letters being less frequent than in the Masoretic Bible Text. Thus î and ú in a medial position are often unmarked (as ʾš Siloam 2 = ʾ ("man"), ṣr Siloam 3; 6 = ṣûr ("rock"). Nevertheless, some new, grammatical material may be derived from them, as the monophthongization of diphthongs even in stressed syllables outside Judah (cf. qṣ Gezer 7, qayiṣ ("summer"); cf. the pun of the prophet Amos (8:2) of this word with qẹṣ ("end"); yn in the Samaria ostraca, yáyin ("wine") or the ending-ô, which is considered by some scholars as dual nominative ending in status constructus (yrḥw ("two months"), Gezer passim), and even the attestation of forms occurring exceptionally in the Bible (as hyt Siloam 3 = hayâṯ ("she was"), a rare form of the third person feminine perfect of verba tertiae yôd, instead of the usual hâyṯâ) is of help. Moreover, they often contain additions to the limited vocabulary of Biblical Hebrew.

Akkadian, Greek, and Latin transcriptions, on the other hand, mark the vowels as well, thus making the recognition of grammatical structure possible. Yet the special conditions of both the transcribed and the transcribing language has to be taken into account, besides the intricacies of transcription itself. Disregard of these pitfalls inevitably results in misinterpretations, as when P. Kahle regarded the double pronunciation of b, g, d, k, p, and t as artificial, inter alia because Origen (185–204 C.E.) and Jerome (342–420 C.E.) deviate in their transcriptions. Kahle did not take into account the fact that Greek and Latin at that time had no means of differentiating between aspirates and spirants (cf. e.g., the use of χ for Arabic k in the Greek papyri of Nessanah, because k was at least slightly aspirated; see Blanc in To Honor R. Jacobson (1967), 298). Moreover, with the exception of Origen's transcription of coherent texts, and words quoted by Jerome, these transcriptions are limited to proper names and thus make an insight into the lingual structure rather difficult. Nevertheless, they are by no means unimportant. Sometimes they exhibit Aramaized forms as against the Hebrew feature according to the Masorah (as Akkadian sa-me-ri-na-ai, Septuagint Σεμερωυ, Σαεμηρώυ; cf. Ezra 4:10, 17 šâmráyin, as against masoretic šômrôn), and vice versa (as masoretic Bnê ḇraq, cf. modern Ibn ibraq, as against Akkadian ba-na-a-a-bar-qa, Septuagint Βαυαιβακάτ [!]). In some cases the noun formation is different (as Akkadian amqa-ru-na, Septuagint ʾάκκάρώύ, exhibiting qaṭṭâlôn (masoretic qiṭṭâlôn), as against masoretic ʿeqrôn, reflecting the parallel qiṭlôn; or masoretic ʾέrεκ corresponding to Septuagint ʾΟπέχ, Akkadian Uruk (also Arku?)). Septuagint ʾΑμμάυ and Akkadian (Bīt-) Ammānu exhibit the stage preceding masoretic ʿammōn (cf. Septuagint variant lecture ʾΑμμών). Even more important are transcriptions exhibiting features preceding the masoretic vocalizations, as Akkadian a-u-si (Hosea) presumably still containing the diphthong, or Akkadian ḫa-za-ki-a-aa, which, as against masoretic Ḥizqiyyâ, perhaps still exhibits the preservation of a in the second and first syllables (cf. also Septuagint ʾΕζεκίας). Since Greek may distinguish between long and short e/o by using η, ω and ε, ο, respectively, Greek transcriptions may even be very important for the recognition of grammatical structures (forms like Septuagint Ησαν = masoretic ʿêśâw, exhibit, it seems, the oldest attestations of the lengthening of vowels in pretonic open syllables). This situation may sometimes be complicated by later sources exhibiting forms that are considered to be earlier, as when the masoretic so-called segolata (as zéḵεr – "remembrance"), similarly transcribed in the Septuagint (as Γαθερ – geṯer), still appear apud Origen as monosyllables (as ζεχρ = zẹkεr).

The Masoretic Text

The Masoretic Text is also not easy to evaluate. It is made up of three historically distinct elements, viz. (in order of their antiquity and stability) the consonantal text, the vowel letters, and the system of diacritical marks for vowels and cantillation.

In the course of time, even the consonantal text underwent changes – an altering in pronunciation led to a change in spelling (since śîn was pronounced as sameḵ, the latter was rarely substituted for it, as stâw ("winter"), Song of Songs 2:11 instead of the original ‡śtâw). Similarly, antiquated forms were replaced by more usual ones (cf. e.g., II Sam. 22:37, 40, 48 taḥatẹnî ("under me"), as against taḥtây, Ps. 18:37, 40, 48, representing the more usual and, presumably, later form), and synonyms replaced obsolete words (cf. parallel passages occurring, e.g., in Chronicles as against Kings, or readings exhibited by "vulgar" Qumran biblical texts). It is even possible that puristic redactions expurgated usages still alive, in favor of literarily preferred features (as Bergstraesser, in: ZATW, 29 (1909), 40ff. assumed for šε being superseded by ʾǎšεr). On the other hand, the differences between Masoretic manuscripts (in contradistinction to the Samaritan version) are so few that this uniformity has to be explained according to the "one-recension" (or even the "archetype") theory.

More conspicuous are differences in the usage of the vowel letters ʾ, h, w, and y: in the Masoretic Text, as a rule, but not always, etymologically long vowels, with the exception of medial â, are marked by one of these letters. The spelling, however, was more defective than in available texts (cf. also the spelling of inscriptions discussed supra), at least at the time of the Septuagint, and there are some differences between the Tiberian and other traditions. Nevertheless, the uniformity mentioned above obtains, as a rule, also in the sphere of the vowel letters. However it has to be assumed that changes affected vowel letters (just as the consonantal text) as obsolete forms were superseded by later ones, and this is exhibited by variant readings occurring in "vulgar" biblical passages found in Qumran. In the Masoretic Text this development is reflected by the so-called ktîv, ("what is written") and qrê ("what is to be read"), two variant readings of which the ktîv, occurring unvocalized in the text, is rejected in favor of the qrê, adduced vocalized on the margin (in many Bibles, however, the ktîv is adduced in the text with the qrê's vocalization, thus causing confusion). Sometimes the ktîv exhibits an older feature, given up in favor of the later qrê (and in many manuscripts what is adduced as qrê in other manuscripts, has already penetrated into the text as a single reading). Thus, it seems that the archaic perfect third person plural feminine -h, to be read , as exhibited by ktîv, was superseded by the ending in qrê. Moreover, the pronominal suffix of the third person singular masculine, after nouns terminating in a consonant, is still sometimes archaically spelled with h in ktîv, as against the more usual spelling with w in qrê. (Sometimes, however, the spelling with -h is the only spelling transmitted.) In other cases the ktîv exhibits the later feature. Thus ʾty, representing ʾattî ("you") (feminine singular), presumably due to Aramaic influence (cf. the Samaritan version and DSIa), sometimes occurs as ktiv, the qrê being ʾatt. The same applies to the perfect second person singular ending -ty, to be read -tî, which was superseded by -t.

The latest stage is exhibited by the vowel and cantillation marks, which developed between c. 600 C.E. (the date of the final redaction of the Talmud, in which they did not yet occur) and the beginning of the 10th century (from which period dated manuscripts have been discovered), but is based on a much older tradition. The only vocalization and cantillation system in use is the so-called Tiberian vocalization. It represents the most elaborate system and is the only one completely preserved. Therefore, it serves as the main base for the grammatical investigation of biblical Hebrew. In principle, however, the other vocalization systems are equally important, i.e., the Babylonian system, which includes several sub-species, and the so-called Palestinian. One has also to take into consideration the Samaritan tradition of pronunciation, and important linguistic features may also be elicited from the Dead Sea Scrolls. The most important innovation, differentiating Tiberian (and one subsystem of Palestinian) vocalization from the others is change of â to å, thus coinciding with å < u. (This feature seems to be very late, however, not after Jerome's time.) There are also other divergences, such as less attenuation of a to i in closed unstressed syllables according to the Babylonian vocalization (as šäḇʿâ = Tiberian šIḇʿâ, which, obiter dictum, already penetrated into Babylonian), a wider supersession of the perfect paʿil by paʿäl, and the preservation of ä in the perfect, imperfect, imperative, and infinitive of hitpaʿel. Yet even the Tiberian system exhibits inconsistencies, and it is difficult to establish whether they are due to the mixture of readings of different subschools (cf. those of *Ben Asher, whose readings have been accepted, and *Ben Naphtali), to chance, or to the desire to be over-accurate. Yet, in spite of all these difficulties in proper linguistic evaluation, the main features of biblical grammar are quite clear.

Despite the multilayered character of the linguistic tradition, the Bible, though stretching over many hundreds of years and emanating from different parts of Palestine, exhibits a surprisingly uniform language. This is due to its being a standardized literary language, on the one hand, and the later changes the text underwent (see supra) on the other. Nothing is known from the Bible even about dialectal differences, with the exception of the fact that the Ephraimites pronounced sibbólεṯ, rather than šibbólεṯ (cf. also supra for more far-reaching monophthongization outside Judah). Post-Exilic books, however, exhibit certain special features which are also found in Middle Hebrew (and sometimes in Aramaic), such as the prevalent use of ʾǎni for ʾânôḵî, and of ʾet with pronominal suffixes rather than their direct annexation to the verb, and the usage of the participle becomes more frequent. Moreover, these books evince a penchant for scriptio plena. On the other hand, poetry, in contradistinction to prose, exhibits certain peculiarities, as the longer forms of the prepositions ʾelê, ‡ʿalẹ, ʿadẹ ("to, on, till") as against prosaic ʾel ʿal ʿad respectively, the less frequent use of the definite article, of the object marker ʾet and the relative pronoun, further the use of endings -î /-ô in the noun in status constructus ( in additional cases as well), the pronominal suffix -mô ("their/them"), the use of status constructus when preceding prepositions, and the extended use of the shortened imperfect.

CONSONANTS

Biblical Hebrew uses the following letters to mark consonants (four of which may also serve as vowel letters, v. supra): א = ʾ, ב = b, ג = g, ד = d, ה = h, ו = w, ז = z, ח = , ט = , י = y, כ = k, ל = l, מ = m, נ = n, ס = s, ע = ʿ, פ = p, צ = , ק = q, ר = r, ׁש (or ׂש) = š (ś), ת = t. The letters k, m, n, ṣ, and p have special forms in final position. The inventory of Hebrew consonantal phonemes, as marked by these letters, is reduced as against the Proto-Semitic one, Hebrew z representing Proto-Semitic z and δ, Hebrew Proto-Semitic and , Hebrew Proto-Semitic ṣ, ḍ, and , Hebrew ʿ Proto-Semitic ʿand ġ, and Hebrew š Proto-Semitic š and . On the other hand, some of the letters in the early period might have been polyphonic, as no doubt was ש, marking, in the Judean dialect at least, both š and ś (differentiated in Tiberian vocalization, e.g., by a point above the letter on its right or left side respectively). It has been claimed that ח and ע were polyphonic as well, exhibiting ḥ /ḫ and ʿ respectively, and that reminiscences of this feature were still alive at the time of the Septuagint, which transcribes ח and ע by zero /χ and zero /γ respectively, χ and γ roughly corresponding to and ġ respectively. On the other hand, these transcriptions may be due to difficulties of transcribing sounds lacking in Greek (cf., mutatis mutandis, the transcription of Arabic ʿ by Greek γ; see Violet, in: OLZ, 4 (1901), 384ff., and the papyri of Nessana, transcribing Arabic and as a rule by zero, once Χαλέδ, Χομαης] by χ; ġ by γ, but once by zero [ʾΑζαλής]. B, g, d, k, p, and t after vowels developed into spirants (even in juncture; differentiated in Tiberian vocalization by the absence of a point, the so-called dageš lene, in them, as against the point in stops), thus entailing the polyphonic use of them in writing. Yet, at first at least, these stops and spirants did not represent different phonemes, and even if they did so later (cf. perhaps ʾalp̱ē – "thousands," status constructus as against ‡ʾalpē – "2,000" status constructus), their phonemical load was very small: this, however, depends on the moot question of the phonemic status of šewa, v. infra. It was in the later biblical Hebrew as well that ś coalesced with s, and this entailed rare cases of mixing them up in spelling (cf. supra). Other important sound shifts, affecting consonants in certain positions, are initial w very early becoming y (with the notable exception of w – "and"), which then often analogically penetrates into medial position as well; the dropping of intervocalic h (as ‡sûsahu becoming sûsô); and w and y dropping in many positions (as ‡yiʿnawu > yaʿǎnε, – "he will be humble," ‡galaya > gala – "he went into exile") and ʾ in some positions (ʾaʾkilu > ʾôḵẹl [pause] – "I shall eat"). N preceding a consonant is assimilated to it. At a later phase, the laryngals and pharyngals became weakened and were no longer apt to be doubled (this applies to r as well); as compensation, the vowel preceding them is often lengthened in this situation. Moreover, әfollowing them appears in the Tiberian vocalization as ǎ, ę ε, and å, and these sounds usually develop after pharyngals and laryngals not followed by a vowel in the middle of the word. Moreover, these consonants often change i / u into a (i at leastinto ε). Doubling of final consonants is given up (as ‡sall > sal ("basket")) (with the partial exception of tt occurring e.g., in ʾatt ("you," feminine singular)), and this occurs even in medial position when followed by a consonant (as tâsoḇnâ ("they will turn," feminine)), and also when the originally following ә has become zero, (as ‡wayyәhî > wayhî ("it was")).

VOWELS

The vowels according to the various vocalization systems differ: mention has already been made that in Tiberian vocalization (and in one subsystem of Palestinian) â shifted to å. Babylonian vocalization does not differentiate between a and ε. Whereas the consonantal script is phonemic (and in some cases even polyphonemic), the vowel marks, especially according to the Tiberian system, can designate auxiliary vowels as well (cf., e.g., the so-called furtive pattaḥ, automatically developing before final h, ḥ, and ʿ after vowels other than a and å). On the other hand, the Tiberian vocalization is polyphonic as well, since absence of vowel, and ә are marked by same sign -ְ, šewa (but in other vocalization systems absence of vowel is not marked at all, this being also the case with final letters in the Tiberian system, with the exception of k, t as stop, and final consonant clusters); it is perhaps the most important moot question of Hebrew vocalization, whether -ְ has or has not to be analyzed as a phoneme, since, e.g., the phonemic status of spirantic b, g, d, k, p, and t largely depends on it (cf. supra ‡ʾalp̱ê as against ʾalpê; if -ְ has to be accorded a phonemic static, ʾalp̱ê has presumably to be phonemicized as ʾalәp̱ê, and then the spirantization would merely result from the preceding vowel). At any rate, historical ә may develop into zero, and vice versa. The Tiberian vocalization (and at least a part of the others) denotes (with the exception of šewa and its allophones ǎ, ě, and ẹ) quality rather than quantity: this is demonstrated mainly by the use of the same sign (-ָ= å) to mark both historical â and u, as well as by the parallel occurrence of a/ε, accounted to be short, and ẹ/o, regarded as long, in certain paradigms (as the verbal paradigms ʿal, pâʿẹl, pâʿol, the nominal paradigms qáṭal ( being h, ḥ, or ʿ), qέṭέl as against qẹ'ṭel, qóṭel; for the usage of pʿl, qṭl v. infra); therefore, length as marked in this article rests on historical reconstruction of a linguistic stage preceding that of the Tiberian vocalization (another change as against Tiberian vocalization, as used in this article, is that, as a rule, vowels were transcribed in accordance with the Sephardi pronunciation, -ָ, as a rule, transliterated by â rather than by å).

The vowels of the Tiberian system (with the exception of šewa and its allophones mentioned above) are (the Tiberian vowel signs, are, as a rule, sublinear, whereas the other systems use superlinear vocalization): ־ַ= a, ־ָ= å, ־ֶ= ε, ־ֵ= , ־ִ = i, ־ֹ or ֹו = o, ־ֻ or ּו = u. Since no quantitative distinctions exist, there is no difference whether or not a vowel sign is followed by a vowel letter (as between ־ִ and ־ִי ;־ֵ and ־ֵי, or ־ֹ and וֹ respectively) and even ־ֻ and ּו are identical: their respective use depends only on whether or not the consonantal text exhibited vowel letters. Whereas the inventory of Hebrew consonants is restricted against the Proto-Semitic one (see supra), that of the Hebrew vowels is extended: Proto-Semitic had, it seems, a system of three short (a:i:u, at an earlier stage presumably a:i /u) and three long vowel phonemes (â:î:û) only, which, of course, were differently actualized. Tiberian a mainly stems from Proto-Semitic a, and also from i in closed stressed syllables (lex Philippi, as ‡ḥafiṣtâ > ḥafaṣtâ ("you wanted")), Tiberian å from Proto-Semitic a, â, and u (in the last case, pronounced according to Sephardi pronunciation o, as against â in the other cases), Tiberian ε from Proto-Semitic a, i, and ay (when preceding å), Tiberian from Proto-Semitic i and ay, Tiberian i from Proto-Semitic i, î, and a (in unstressed, closed syllables, see supra), Tiberian o from Proto-Semitic u, â (in formerly stressed syllables, then often analogically spreading to other positions as well), and aw, and Tiberian u from Proto-Semitic u and û. As shown, the diphthongs aw and ay were monophthongized, becoming o and e respectively. Sometimes, they are preserved, as normally before double w and y, e.g., gawwḵâ ("your back"), ḥayyîm ("life"), and in other conditions, like mnūḥayḵî ("your rest," Ps. 116:7) and ʿawlâ ("wickedness"). In closed syllables bearing the main stress aw and ay were split into two syllables, as ‡tawk = táωεḵ ("middle"), ‡bayt >< báyiṯ ("house"), iy often becomes î, especially after prepositions (as ‡liyhûḏâ > ‡lîhûḏâ ("to Judah"); so also ‡miyyәḏê > ‡miydê > < mîḏê ("from the hands of…"). In Babylonian vocalization , especially in initial position, is graphically represented by yi, which, according to some scholars, is intended for the pronunciation of ī (as yirâ ("be afraid!"), as against yәrâ), and similarly in Tiberian vocalization w- ("and") before labials, and consonants followed by ә, becomes u (as umέlεḵ ("and king")). Moreover, according to the Ben Naphtali school of the Tiberian system, medial әyi is apt to shift to i. ә (marked by šewa) developed from original short vowels in open syllables two or four syllables before the stress (as ‡ladabaraykúmu > lәḏiḇәrẹḵέm later lәḏIḇrẹḵέm, ("to your things"); sometimes short vowels in open pretonic syllables are reduced (as ‡masmẹrîm > masmrîm ("nails"); so always in status constructus and prepositions, bearing only secondary stress or no stress at all, as ‡dabar- > dәḇar-); in other cases again the following consonant is doubled, thus enabling the retention of the short vowel, as qṭannâ ("small," feminine); as a rule, however, pretonic short vowels (especially a) in open syllables are lengthened (as ‡dabar > dâḇâr ("thing"), ‡ʿinab> ʿẹnâḇ ("grape"), as well as in open stressed final syllables. In closed stressed final syllables vowels are lengthened, as a rule, in nouns in status absolutus (as yâḏ ("hand")), but not in status constructus (as yaḏ-) and in verbs (as kâtaḇ ("he wrote")), presumably because the latter lost their short final letters earlier, thus terminating in closed syllables. Eventually, however, short final vowels were generally omitted.

STRESS

The stress is transmitted by the cantillation marks: it falls on the last or the penultimate syllable. Because oxytones are more frequent, in this article, as a rule, only paroxytones are marked as such. Since, as a rule, the penultimate syllable is stressed when no final syllable was lost, whereas the last syllable bears the stress, when the word terminated in a short final vowel at an earlier stage, a stage of general paroxytone accent may be reconstructed. The few exceptions, as kâṯḇ'â ("she wrote"), kâṯḇû ("they wrote"), and similar perfect, imperfect, and imperative forms, further e.g., yâḏḵ'ā ("your (masculine singular) hand"), also exhibiting exceptional reduction of the vowel in the syllable preceding the stress (see supra), have to be regarded as later forms, the original syllable pattern being preserved in the pausal forms kâṯ'âḇâ, kâṯ'âḇû, and yâḏ'êḵâ. An even later stage of stress is exhibited by the so-called segolata, in which the cluster was split at a very late stage (as ‡sipr> sepεr ("book"); cf. supra for the transcription of these forms). An early stage of Proto-Hebrew stress may tentatively be reconstructed by the assumption (cf. supra) that it was stressed â that shifted to ô.

PRONOUNS

Hebrew (as Semitic) word formation exhibits, as a rule, tri-radical structure: the main meaning is carried by the (generally three) radical consonants, while the vowels only add shades to it. Yet particles (as much as not of nominal origin) and pronouns deviate from this structure, pronouns also allowing word composition, a feature alien to Semitic linguistic system (as hallâzě ("this"), compounded from the demonstrative that serves as definite article as well+lâ+zě).

"I" is expressed by ʾǎnî and ʾânôḵî, both occurring in pause as ʾ'ânî / ʾân'ôḵî, exhibiting a more original stress structure (as in the case in pause in most cases, cf. supra; for ʾân'ôḵî being the original form; cf. also the preservation of â, as well as ô < ‡â peculiar to stressed syllables, v. supra). The same is the case for pausal ʾ'âttâ (also ʾ'attâ) as against the context form ʾattâ ("you," masculine singular). In some rare cases ʾattâ is spelled defectively without the usual final h, i.e., ʾt (e.g., I Sam. 24:18), always as ktîv, in others again ʾt is vocalized ʾatt (e.g., Num. 11:15). For ʾ'áttî (ktiv) occurring for ʾatt ("you," feminine singular), v. supra. For hi (ʾ) ("she"), (ʾ) ("he") is substituted as ktîv in the Pentateuch, for which, as qrê perpetuum, hî (ʾ) is read. הואה ,היאה ("he, she"), as occurring in the Qumran scrolls, exhibiting long forms with final vowels, are, it seems, due to secondary analogical formation. For the regular ʾǎnáḥnû ("we") náḥnû occurs rarely, and once (Jer. 42:6) the ktîv ʾnw, identical, it seems, with mishnaic ʾ'ânû. Whereas hem and hémmâ (Babylonian häm and hämmâ; "they"), ʾattẹn (once, Ezek. 13:20) and ʾatt'ẹna (Babylonian ʾattän and ʾatt'ännâ, "you," feminine plural) alternate, he'nna (Babylonian hẹ'nnâ; "they," feminine plural) is the only existing form, rather than ‡hẹn. ʾattεm ("you," masculine plural) is the only existing form in the masoretic text, ʾtmhattémâ or perhaps ʾattẹ'mmâ) occurring only in "vulgar" versions found in Qumrân, presumably exhibiting (as does hẹmmâ) a late analogical formation. For marking possession, etc., and direct object, identical pronominal suffixes attached to nouns, prepositions and verbs respectively are used in Hebrew (and in Proto-Semitic), the only difference between them being in first person singular, -î denoting "my" (and, as a rule, attached to prepositions) and -nî "me." If the word to which the pronominal suffixes are attached terminates in a consonant, it is, as a rule, preceded by a "connecting" vowel, a being the favorite vowel after perfect, after imperfect and imperative. Yet before -ḵεm /ḵεn the "connecting" vowel is missing (as yεḏḵεm ("your hand") as against, e.g., yâḏẹ'nû ("our hand")). The absence of the "connecting" vowel before the pronominal suffix of the second person singular masculine (as yâḏḵâ ("your hand")), however, is secondary, see supra. On the other hand, pausal forms such as lâḵ ("to you," masculine singular) display the omission of the final syllable. -ḵâ is sometimes spelled with a final h, and this spelling becomes more frequent in the Qumran scrolls. After nouns, prepositions, and imperfect forms terminating in consonants the pronominal suffix of the second person feminine singular is -ẹḵ, rarely -âḵ (kullâḵ ("you in your entirety"), presumably influenced by lâḵ ("to you" where the â belongs to the particle) or -ẹḵî. -ô ("his"), spelled archaically also with -h, v. supra, is sometimes superseded by -ẹḥû-, which is the regular suffix only after imperfect, imperative, and nouns terminating in ε, but becomes more frequent in the Qumran scrolls. After perfect forms, and -âhû alternate. Tag̱múl'ôhî ("his benefits," Ps. 116:12), instead of ordinary tag̱mûlâw (-âw being the usual suffix after plurals), exhibits Aramaic influence. For -âh ("her"), also used after perfect (the consonantal value of the h being marked by a point in it, the so-called mappîq), quite often occurs (without mappîq, the h used as vowel letter only). For -ḵεn (second person plural feminine), -hεm, -hεn (third person plural masculine and feminine respectively), forms with final rarely occur: -ḵεnâ, -hěmâ, and hεnâ respectively, and even kεm (second person plural masculine) is attested in the Qumran scrolls with final-â, where this ending is especially frequent. For-hεm the archaic poetical - occurs. For-ân (which together with-âm replaces -hεn and-hεm respectively after nouns and perfect forms terminating in consonants, whereas-êm is used after imperfect forms for the masculine) after nouns 'ânâ is rarely attested. In nouns terminating in the plural ending-ôṯ, for ôṯẹhεm /-ôṯẹhεn ("their," masculine/feminine) -ôṯâm / -ôṯân occurs. Suffixes after imperfect and imperative forms, as well as after certain particles, are sometimes preceded by -εn, the so-called nûn energicum, as yḇârǎḵεnhû ("he will bless him"), generally assimilated to the following consonant, as yakkεkkâ, ("he will smite you"), or assimilating a following h, as yišmrεnnû ("he will keep him").

The demonstrative pronouns (masculine singular), zôṯ (feminine singular, rarely ),ʾ'ẹll ε (plural, rarely ʾẹl) ("this"), exhibiting the alternation of δ -root in singular and l -root in plural, well known from other Semitic languages, are, if used attributively, preceded by the definite article, in analogy to the usage of attributive adjectives (as hayyôm hazzε, "this day"). This construction is to be regarded as later than that occurring in Middle Hebrew, exhibiting both the noun and the attributive demonstrative pronoun without the definite article (as yôm zε). Accordingly, a phrase like ballaylâ hû (instead of the regular ballaylâ hahû, exhibiting the use of the third person of the personal pronoun as "that" – demonstrative), in which the definite article is attached to the noun, rather than to the demonstrative, is, it seems, the intermediary stage. The definite article ha- (with doubling of the following consonant) has still retained its demonstrative force in phrases like hayyôm ("this day, today"). On the other hand, is sometimes used as presentative, or as relative pronoun (as is also ). Mâ/mε ("what"), spelled mh, with doubling of the following consonant (this doubling is a real one, not like that called dḥîq or ʾâṯê mẹraḥîq), has presumably, partly at least, to be derived from ‡mah/ ‡mεh. "Who" is .

TENSE SYSTEM

The Hebrew tense system, besides the imperative (in second person only, in form closely related to the imperfect), consists of four finite forms, viz. the perfect and the consecutive (the so-called conversive) perfect, the imperfect and the consecutive (conversive) imperfect. The consecutive tenses are preceded by w- ("and"), which before the imperfect has the basic form wa (with doubling of the following consonant). It is a moot question whether this system marks aspects (without any notion of time) or rather time. At any rate, in biblical prose at least, these forms seem to denote time, since the difference between perfect and consecutive imperfect, referring to the past, and imperfect and consecutive perfect, referring to the future/present respectively, depends, it would seem, on the syntactical environment only: as a rule, whenever it makes the use of w-, etc. ("and") possible, the consecutive forms are used, in accordance with the demanded time, otherwise simple imperfect/perfect are applied. Besides the indicative, the imperfect has a cohortative (especially in the first person), formed by the ending , (also occurring after consecutive imperfect and the singular masculine of the imperative), and a jussive. The latter, though often coinciding with the indicative, even more resembles the consecutive imperfect, both being formed from the apocopate; the main difference between them is the paroxyton stress of the consecutive form (presumably an archaic feature, reflecting the stage in which, see supra, general paroxyton stress obtained, the indicative being ‡yaf ʿálu > yaf ʿál, the apocopate ‡yáf ʿal. The jussive was then more fully adjusted to the stress pattern of the indicative than the consecutive imperfect). In verba tertiae y, apocopate forms of the imperative occur as well. As to the consecutive perfect, it often exhibits oxyton stress, as against the paroxyton stress of the perfect. Yet this oxyton stress is, it seems, secondary, since syllable structure is in accordance with the paroxyton stress (as ʾâḵalt'â ("and you shall eat"), parallel to ʾâḵáltâ ("you ate"), rather than ‡waʾǎḵalt'â).

The perfect is formed by afformatives, which in the first and second persons resemble the endings of the personal pronouns (but first person singular terminates in -). The third person singular masculine has the ending zero, the feminine -â, the plural-û.

Less clear are the formatives of the imperfect, which uses the prefixes ʾv (v marking any vowel) for first person singular, tv for second person, and third person feminine, yv for third person masculine, and nv for first person plural. Other persons are indicated, apart from the mentioned prefixes, by suffixes as well: (rarely -în) in second person feminine singular (as tišmʿî "you will hear"), (rarely -ûn, which, very seldom, penetrates into the third person plural of the perfect as well) in second and third person masculine plural (as tišmʿû /yišmʿû ("you/they will hear")), and - in second and third person feminine plural (as tišmaʿ ("you/they (feminine) will hear"); rare is the yv prefix in third person feminine, as yʿaămoḏnâ ("they (feminine) will stand," Dan. 8:22)). The imperative has no prefixes, but the same suffixes; for- very rarely ‡-n occurs: šmáʿan (< ‡šmáʿn) ("hear (feminine plural)!," Gen. 4:23) (in most cases forms of the imperfect and imperative terminating in n are considered as scriptio defectiva, and vocalized -).

Among infinite forms, the Hebrew verbal system possesses a participle, which may behave as a noun (it may, e.g., stand in status constructus and be negated by ʾ'ẹn, the negation of nominal sentences), on the one hand, and as a finite verbal form, on the other (it may, e.g., govern direct object and be continued by a consecutive finite form, as makkẹ ʾiš wâmẹṯ ("if one smites a man and he dies"), although the participle stands in status constructus). Among the two infinitive forms, the so-called infinitivus constructus has usual infinitive functions, as mẹṭîḇ naggẹn ("he who excels in playing," Ezek. 33:32); wayyẹrεḏ… lirʾôṯ ("and He descended… to see," Gen. 11:5). Very often the l- form is used even when not in final sense, exhibiting the coalescence of l with the infinitive, as ḥâfẹṣ… lahămîṯẹnû ("He wanted… to kill us," Judg. 13:23; in Middle Hebrew this form becomes the only existent infinitive form). It is also used as gerund, as bârâ… laʿăśôṯ ("he created… in making," Gen. 2:3). The so-called infinitivus absolutus, so called, because it does not stand in status constructus nor is it governed by preposition, is a peculiar blend between verbal noun and verbal interjection. It is, besides its rare infinitive functions, mainly used as internal object (as bêraḵtâ ḇârêḵ ("surely you have blessed," Num. 23:11), as a rule preceding the finite verb (as môṯ yûmâṯ ("surely he will be killed," Gen. 26:11); also replacing modal adverbs (as wayy'ẹlεḵ hâlôḵ wәʾâḵôl ("he went while eating," Judg. 14:9), and as substitution of finite verbal forms (as zraʿtεm harbẹ whâbẹ mʿâṭ ʾâḵól wʾẹn-lśoḇʿâ, "you have sown much, but brought home few, you have eaten, but not to satiety," Hag. 1:6), mainly of the imperative (as zâḵôr ʾεt-yôm haššabbâṯ ("be mindful of Sabbath")).

VERB PATTERNS

Of all word classes it is the verb that has the most conspicuous patterns, although patterns as such are one of the main characteristics of Semitic languages in general and of Hebrew in particular. These patterns are characterized by a certain vowel sequence, which, interwoven with the trilateral root, together with the repetition or doubling of radical consonants, as well as the addition of certain formative consonants, reflects various modifications of the root connected with specific meanings. It is customary to denote the verbal themes by the root pʿl (which, however, has the disadvantage of not being able to denote the doubling of the second radical, since ʿ cannot be doubled) in accordance with the vocalization of the third person masculine singular of the perfect. The usual verbal patterns in biblical Hebrew are the ground theme ʿal (also called qal); its reflexive-passive nif ʿal; piʿẹl, exhibiting doubling of the second radical and denoting intensive and factitive action, its passive puʿal and reflexive-reciprocal theme hitpʿael, the causative hif ʿîl and its passive hof ʿal. Beyond these themes a stage may be reconstructed which, with the exclusion of nif ʿal, exhibits a well-balanced system:

Hebrew Language

Ground
pattern
Double
pattern
Causative
pattern
ʿal pʿiel hif ʿîl
internal passive t-form (reflexive reciprocal) puʿal puʿal hof ʿal
hiṯpâʿal hiṯpaʿẹl ‡hiṯaf ʿẹl

From the patterns marked by a double dagger (‡) only remnants exist. The inner passive of paʿal disappeared because of its resemblance to the inner passive of perfect piʿẹl and imperfect hif ʿîl, being superseded by nif ʿal, which, besides its original reflexive meaning, acquired passive functions as well; it can only be recognized by its perfect being identical with puʿal, the passive of piʿẹl, its imperfect with hof ʿal, without a corresponding piʿẹl /hif ʿîl. Hiṯpáʿal subsisted in hiṯpâqḏû ("they were counted") only (which is mixed up with its passive hoṯpâqḏu) whereas the very existence of hiṯaf ʿẹl is dubious, depending on the analysis of forms like tәṯaḥărê ("you will compete"), wattẹtaṣṣab ("and she stood").

In ʿal, the neutral perfect forms ʿẹl and especially ʿol are being superseded by ʿal. In the imperfect and imperative yaf ʿol /pʿol and yaf ʿal /pʿal respectively are alive, yaf ʿil /pʿil being absorbed by hif ʿil and not really subsisting but in some "weak" roots. From vestiges a stage may be reconstructed exhibiting the imperfect forms yaf ʿol (and yaf ʿil) as against yif ʿal, cf. yεḥεlaš as against yaḥăloš; yêḇôš as against yâqûm; yêqal as against yâsoḇ. The only living infinitivus constructus is pʿol, pʿal (as liškaḇ ("to lie"); the stop k, as against in šḵaḇ, is due to the coalescence of l – with the infinitive, v.supra) being marginal and pʿil existing in some "weak" verbs only. Feminine forms of the infinitivus constructus are attested as well, as ʾahăḇâ ("to love"). The infinitivus absolutus has the form ʿôl, ô in the second syllable also occurring in other themes. The active participle, originally belonging to ʿal, is ʿẹl, whereas ʿẹl and ʿôl are the original participles of ʿẹl and ʿol respectively, yet losing ground against ʿel. The passive participle is ʿûl (whereas the participle of the internal passive of ʿal is puʿâl and piʿol, with redoubling of the second radical, as ʾukkâl ("eaten"), yillôḏ ("born")).

Nif ʿal, exhibiting attenuated i in its first syllable, still preserves a in some "weak" roots. In the imperfect, the imperative and the infinitive (the two latter forms have a hi- prefix), the n, immediately preceding the first radical, is assimilated to it; the last radical of these forms is, as a rule, preceded by Tiberian , but Babylonian ä. In contradistinction to all the other themes, with the exception of the ground theme, the participle is not formed with the prefix m-, but in accordance with the perfect (nif ʿâl).

Piʿel exhibits a in perfect in the second syllable about as often as ; in the Babylonian vocalization ä (corresponding to a /ε) prevails, and some few verbs (as dibbεr, "he spoke") exhibit ε according to Tiberian vocalization as well. In the imperfect, etc., the first radical is followed by a, the second by e in the Babylonian system sometimes by ä).

Hiṯpaʿẹl sometimes exhibits a in its third syllable in all forms (with the exception of the participle), especially in pause, and in the Babylonian vocalization the corresponding ä prevails. If the first radical is a sibilant, it precedes the t (as hištappẹk ("to be poured")). Sometimes, the t is assimilated to the first radical (as hiṭṭammẹ /hiṣṭaddẹq ("to be defiled/to clear himself")).

In hif ʿîl the vowel of the prefix is i in the perfect (a occurring in "weak" verbs only), a in the other forms; that of the second syllable, as a rule, stressed î, before consonantal afformatives in the perfect a, in the imperfect, etc. , which is exhibited also by infinitivus absolutus and those forms of the imperative and jussive which lack conjugation endings and pronominal suffixes. The h of the prefix is, as a rule, dropped in the imperfect.

Pu'al and hof ʿal have the vowel sequence u /o-a, characteristic of the internal passive.

WEAK VERBS

Verbs are that class of words in which triliteralism is most strictly carried out. Nevertheless, some verbal classes, viz. mediae infirmae and geminatae, cannot be explained on the assumption of sound shifts operating on triliteral roots only: they have to be considered as partly emerging from biliteral roots, blending with forms of triliteral roots, which underwent changes because of "weak" letters; and this may apply to other "weak" verb classes as well.

Verba primae n assimilate the n to the immediately following second radical, this being in accordance with the general behavior of n (see supra). Moreover, those having the imperative pʿal in the ground theme, drop the n in it (as gaš ("approach")), as well as in the infinitivus constructus, which terminates in the feminine ending -t (as ‡gašt > gεšεṯ). The same is the case with nâṯan ("to give"), which is the only verb primae n having an i – imperfect (yittẹn): tẹn, ‡tint > tẹṯ. It is also the only verb tertiae n which assimilates this n to consonantal suffixes, as nâṯattî (as against e.g., sâḵantî, where the n is analogically restored). It may be due to the influence of nâṯan, that its antonym lâqaḥ ("to take") treats its l like an n: yiqqaḥ, qaḥ, qáḥaṯ. Also some verba primae y (especially those having ṣ as their second radical, as niṣṣab ("he stood") as against hiṯyaṣṣẹb) and mediae infirmae (as massîg̱ ("removing") as against nâsôg) behave like primae n.

Verbae primae y form yaf ʿil imperfect, its imperative and its infinitive omitting the y: yâšaḇ ("to seat"), têšẹḇ (with assimilation of the prefix vowel to the following one), šẹḇ, šεḇeṯ (< ‡šiḇt, ә with feminine ending); yâḏaʿ ("to know"), têḏaʿ(< tәdẹăʿ, cf. the infinitive-noun ʿâ), daʿ, ʿaṯ. Hâlaḵ ("to go") behaves similarly: têlẹḵ, lẹḵ, lεḵεṯ. Yâḵol ("can") has the solitary imperfect yûḵal. Nif ʿal has in perfect and participle ô (nôlaḏ ("was born"), nôlâḏ, exhibiting the prefix vowel a, cf. supra, as does the whole paradigm of hif ʿîl, as hôlîḏ), the other forms, as a rule, exhibit w as first radical. Hof ʿal has û (as hûḇal, "he was brought").

In verba tertiae infirmae, y has superseded w. Forms without suffixes have the same endings in all themes: -â in perfect (and әṯâ in feminine singular third person), in imperfect and participle, - in imperative, -ôṯ in construct infinitive. Consonantal afformatives are preceded in perfect by -ẹ or î in unstable distribution, î only being used in ʿal (as if continuing ʿil); ˆεnâ (ˆε < [< ay due to assimilation to the following â, pronounced å]) is the suffix of imperfect second and third person, and imperative second person, plural feminine. Before vocalic afformatives y and the preceding vowel drop (as ʿâśû- ("they did"), ʿăśî ("do," feminine singular)). In jussive and the consecutive imperfect the second radical syllable is omitted (as ʿâś, "let it be done"); final double consonants thus arising are simplified (as wayman ("and he appointed"); so also in the shortened form of the imperative: ṣaw ("order!")). Final consonant cluster may be preserved, if the second consonant is a stop (as wayyebk ("and he wept")), or as a rule, broken up (as yîrεḇ < ‡yîrb ("let it multiply")).

In verba tertiae ʾ, the ʾ is dropped (preserved as vowel letter only) in final position and a preceding short vowel lengthened (as ‡mâṣaʾ < mâṣâ, ("he found")). Before consonantal afformatives of the perfect ʾ is dropped, â (־ָא) being the preceding vowel in ʿal (as mâṣâṯî) and, it seems, in puʿal and hof ʿal (the only instance being hûḇâṯâ ("you have been brought, Ezek. 40:4)). ê (־ֵא) in the other perfect forms (as yârẹṯî ("I feared," ʿẹl form of the ground theme), niqrẹṯî ("I was called")), as well as in the feminine singular of the participle (as mûṣẹṯ ("being brought out")). The suffix of imperfect second and third person and imperative second person, plural feminine, ends, as in tertiae y, in -ˆεnâ, yet exhibiting ʾas vowel letter, rather than y as in tertiae y.

Verba mediae infirmae exhibiting in the first and second persons of the perfect of the ground theme short vowels (as qamtî, "I stood up"), but long ones in the third (as qâm, corresponding to the ʿẹl- form mẹṯ ("he died") and the (synchronically) ʿol-form bôš ("he was ashamed")). In the (regular, jussive, and consecutive) imperfect, the imperative and infinitivus constructus of the ground theme, verba mediae w exhibit û, etc., as yâqûm, y'âqóm, wayyâqom (with qâmâṣ qâṭân in the last syllable), qûm, qûm (but second/third person feminine plural tâšoḇnâ, along with tәmûṭˆεnâ, exhibiting, as do mediae geminatae, the same ending as tertiae y); mediae y î, etc., as yâśîm ("he will put"), yaś'ẹm, wayy'âśεm (these forms being identical with the corresponding hif ʿîl forms; therefore ground themes of mediae y are apt to pass to hif ʿîl, as bántâ ("you understood") to hẹḇîn), śîm, śîm. The old yif ʿal imperfect is preserved in yẹḇôš. The active participle is identical with the perfect, but in contradiction to it is always oxyton (this applies to the other themes and to mediae geminatae as well), even before vocalic afformatives: perfect qâm, q'âmâ, q'âmû, 'mêṯ, m'êṯâ, m'ēṯû, bôš, b'ôšâ, b'ôšû, as against participle qâm, qâm'â, qâm'îm, qâm'ôṯ, mêṯ'â, bôš'â, etc. The imperfect and imperative too have paroxyton stress before vocalic afformatives, and this applies to perfect, imperfect, and imperative nif ʿal and hif ʿîl and mediae geminatae pâʿal, nif ʿal, hif ʿîl and hof ʿal as well. Perfect and participle nif ʿal is like nâsôg ("to retreat"), hif ʿîl like hêqîm, mêqîm (where the preformative has the same vowel as that of the perfect); the perfect consonantal afformatives of these two themes as well as of ʿal, nif ʿal, and hif ʿîl of mediae geminatae are preceded by ô, as nәsûg̱ôṯî (exhibiting û rather than ô in the radical syllable), hăḇîʾôṯîw ("I brought him," along with hẹḇ'ẹṯî; rarely also forms like wahăqẹmônû with occur). Nif ʿal imperfect, etc., exhibits ô (as yissôg). Hof ʿal (as also that of mediae geminatae) is formed on the analogy of primae y. For piʿẹl, puʿal, and hitpaʿẹl, in classical language polẹl, etc., is used: qômẹm, qômam, hiṯqômẹm (externally identical with the ʿẹl, etc., forms of mediae geminatae, in which (along with forms like sibbẹb, gullal, hitḥammem) forms like sôbẹb, gôlal, hiṯgôlẹl are used for piʿẹl, etc. From both verb groups pilpẹl, etc., themes may be derived, as gilgẹl from gll or ṭiltẹl from ṭwl).

Besides "strong" forms, verba mediae geminatae, as a rule, exhibit forms containing one radical syllable with doubling of the second (= third) radical (which, however, is simplified when not followed by a vowel, as ʿal raḇ ("was much"), nif ʿal timmaqnâ ("they will be consumed")). Along with them forms with reduplication of the first radical occur, the so-called Aramaic formation, together with the doubling of the second radical (as hoššammâ ("its being desolate")), and without it (as wayyiqqdû ("and they bowed")). Forms without any reduplication are also attested (as yâzmû, ("they will plan")). In the imperfect of the ground theme, yaf ʿol (as yâsoḇ, consecutive imperfect wayy'âso¸ḇ) corresponds to a perfect (when without ending or with vocalic afformatives) and participle built according to the "strong" pattern (as sâḇaḇ, sâḇәḇû, sôḇẹḇ), whereas yiʿfal (derived from adjectives exhibiting a in their sole syllable, the final consonant being doubled when followed by afformatives, as yẹqal ("he will be easy") from qal ("easy")) has them in the "one radical syllable" form (as qal, qállâ, qállû, qal, qallâ, qallîm, qallôṯ). From the yaf ʿil imperfect only remnants exist (as yâg̱ẹn, "he will defend"). Perfect/participle nif ʿal have two forms: násaḇ/nâsâḇ (also participle nâg̱ẹl) and nâḇoz/nâḇôz, and the same applies to imperfect/imperative: yiddam and yissoḇ. Hif ʿîl perfect has a (as hêqal) or (as hẹḥẹl; the Babylonian system ä only), the imperfect (as yâqẹl; the Babylonian has sometimes ä), the participle ˆ in both the radical and prefix syllable: mˆẹsˆẹḫ (as have some perfect forms, v. supra).

SUBSTANTIVES

Both triliteralism and the development of patterns is less conspicuous in nouns than in verbs. There exists a set of biradical substantives with a fixed vowel, which, by their meanings, demonstrate that they belong to the oldest stratum of the language: yâḏ ("hand"), dâm ("blood"), dâg ("fish"), bˆẹn ("son"), šˆẹm ("name"), ʿêṣ ("tree"), šânâ ("year"), śâp̱â ("lip," in plural transferred to triradical scheme by the inclusion of the feminine ending: sip̱ṯô'ṯˆεḵâ), etc. The notion of patterns is best developed in verbal nouns, in participles and infinitives, as qiṭṭûl (q, ṭ, l denoting the three radical letters respectively) belonging to piʿẹl, moreover in nouns with m-prefix, especially in nomina instrumenti exhibiting maqtel, less in those with t-prefix. The suffixes include -ôn and -ûṯ (containing the feminine ending as well). Among nouns without affixes qâṭîl, qaṭṭîl are frequent in adjectives, qâṭôl, plural qṭullîm, denotes color adjectives, qiṭṭẹl bodily or mental faults, qaṭṭâl intense qualities and occupations. One-syllabic nouns, terminating in a consonant cluster, open the cluster, mostly by ε (segol, the so-called segolata, see supra), as yεleḏ ("child") ‡yald, ŝεḇεṯ ("to sit") < ‡šibt (in pause yâleḏ; š'âbeṯ, yet mεlεḵ and most nouns having original qiṯl pattern do not change in pause).

Substantives are used in different status:, in status absolutus, when standing alone; in status constructus, when closely attached to a following noun (the so-called nomen rectum, historically a genitive; the nomen rectum defines and, when itself determinate, determines the noun in status constructus); and status pronominalis, when attached to a pronominal suffix, which stands in the same relation to the noun as the nomen rectum does. The feminine ending is either (in status constructus and pronominalis -at, etc., exhibiting an earlier stage), or -t; sometimes these feminine endings alternate, as when used in status absolutus and t in status constructus and pronominalis, as mεrkâḇâ ("chariot"), mirkεḇεṯ- (< ‡mirkaḇt, exhibiting the opening of the final cluster, as in the segolata), mεrkaḇtô. The dual is rather reduced, being as a rule used with "two," "two hundred," some nouns denoting time and mainly with objects which naturally occur in pairs, especially the double members of the body. Its ending is -áyim, that of the masculine plural -îm, the status constructus and pronominalis of both -ˆ (historically to be derived from the dual), which is also added to the status pronominalis of the feminine plural ending -ôt. The so-called segolata, including one-syllabic nouns with feminine ending, as yaldâ ("girl"), form their plural from bisyllabic stems, exhibiting â after the second radical: ylâḏîm ("children"), ylâḏôṯ ("girls"). Mention must also be made of the unstressed locative ending (spelled -h, < ‡ah, as intimated by Ugaritic), as ṣâ'pônâ ("northward"), also occurring between status constructus and its nomen rectum, as midbárâ dammεśεq ("to the wilderness of Damascus").

ADJECTIVES

The boundaries between substantives and adjectives are rather blurred. There are relatively few patterns exclusive to one of these word classes (as segolata mainly for substantives, qâṭôl, plural qṭullîm for color adjectives). Adjectives invariably have feminine forms ending in-â/t, masculine plural terminating in-îm and feminine plural in-ôt; substantives, on the contrary, need not have special feminine forms and also feminines without endings occur (as ʾâṯôn ("she-ass")). Moreover, in the substantives singularia/pluralia tantum occur, and the plural ending-îm may be attached to feminine substantives (as ʾiššâ ("woman"), plural nâšîm), and even more often vice versa (as ʾâḇ ("father"), plural ʾâḇôṯ). Yet some substantives exactly behave as adjectives in this respect (as yεlεḏ, yalḏâ, ylâḏîm, ylâḏôṯ). Adjectives proper do not have status pronominalis, yet substantival usage of adjectives (and sometimes also vice versa) is frequent. Sometimes adjectives may occur in status constructus, yet their usage is very special (as nqî ḵappáyim ("clean as regards hands")), and it may not be substituted by status pronominalis (as if ‡nqîhεn; but this may apply sometimes to substantives in status constructus as well). Some syntactical usages, however, seem to be possible for adjectives only, rather than for substantives, as the use of modifiers like ʾôḏ ("very"), yôṯẹr ("more"). The simplest solution would perhaps be to set up three different classes: substantives, adjectives, and finally nouns, which would then include both word classes, as far as their special character cannot be defined by formal criteria. Adjectives used as attributes are preceded by the governing noun.

NUMERALS

As to the cardinal numbers, ʾεḥâḏ, feminine ʾaḥáṯ ("one") is mainly used as adjective, šnáyim, feminine štáyim (sic! with t as stop, exhibiting a quite exceptional initial cluster) "two," as substantive, governing the counted noun in status constructus: šnẹ-, feminine štẹ-, or following it in status absolutus. As to the numbers three to ten, those with zero-ending refer to feminine nouns, whereas those with -â (in status constructus -at/-t) to masculine ones, a common Semitic feature, in opposition to the other noun classes. They precede or follow the counted nouns in status absolutus, but they may precede them in status constructus as well (historically an archaic feature): this is the rule with definite nouns, as well as with yôm, etc. The "ten" in the numbers 11–19 is ʿâśâr for masculine, ʿεśrẹ for feminine, spelled ʿśrh; it is uncertain whether it exhibits an alleged feminine ending -, since in Ugaritic it is spelled with final h, thus intimating a consonantal ending. The ordinal numbers have special forms only from one to ten, exhibiting the theme qṭîlî, with the exception of šiššî ("sixth"), and perhaps šênî ("second"). "First" is rîšôn, a relatively late form, as it is customary in Semitic language, the older usage being the use of the cardinal number "one," still persisting in biblical Hebrew.

PREPOSITIONS

Prepositions, as far as etymologically transparent, are as a rule nouns in status constructus/pronominalis, as ʾẹṣεl ("near, by"), ʾeṣlkâ; ʿal-, poetical ʿălẹ ("on"), ʿălẹḵεm, the suffix -ẹḵεm originally exhibiting the final -y root, rather than a plural suffix. Through the influence of prepositions like ʿălεkεm, other prepositions too govern plural suffixes, as táḥat ("under"), taḥtệḵεm. Among the three uni-consonantal prepositions, only l- ("to"), and b- ("in, by") govern pronominal suffixes (as lḵâ, lâḵ, lâḵεm), but k-, originally a demonstrative element, governs pronominal suffixes mainly by means of --: kâm'ôḵá, kmôḵεm. Min ("from") as a rule assimilates its n to the following consonant (as mibbnô ("from his son"), mikkεm, cf. also mẹhεm, with lengthening of the vowel preceding the pharyngal not capable of doubling); before some suffixes min is doubled, as mimmέnnî < ‡minminî ("from me"), along with poetical minnî. There are two prepositions ʾẹṯ; one denoting "with," has the form ʾittô, etc., before pronominal suffixes, the other, used as optional mark of determinate direct objects, the forms ʾôṯô, ʾeṯḵέm; sometimes, however, these two sets become mixed up. Since the impersonal passive may govern objects (as indirect object, e.g., yḇullaʿ lammέlέḵ ("the king will be afflicted," II Sam. 17:16)) or a direct one, e.g., maṣṣôṯ yẹʾâḵẹl ("unleavened bread shall be eaten," Ex. 13:7), it may govern ʾẹṯ preceding the definite object as well, e.g., ʾέṯ-kol-dg̱ê hayyăm yêʾâsẹp̱ lâhεm ("shall all the fishes of the seas be collected for them?" Num. 11:22).

NEGATION

As word negation and in verbal clauses is used, in nominal clauses ʾẹn, in prohibition ʾal with the imperfect (jussive; but, as in Semitic languages in general, never with the imperative).

CLAUSE FORMATION

It is in the domain of clause formation that Hebrew has best preserved the ancient Semitic character. In contradistinction to Arabic, it has not relinquished free sentence structure in favor of systematization. Yet, although it has lost, like Aramaic, the case and mood endings, it has not been affected by a similar syntactic formlessness. The boundary lines between main and subordinate clauses are blurred, since w- ("and") may precede the main clause following the subordinate one; cf. also w- introducing the main clause after phrases like wayhî ʾim…, w-… ("and it happened, when…, then…"). Circumstantial clauses resemble main clauses even more, mainly differentiated by the use of different tenses. Moreover, the number of subordinate conjunctions is relatively small, the most important ones being the relative pronoun ʾăšεr, also used as introducing substantive clauses, introducing substantive and causal ones, conditional ʾim and hypothetical . Very frequent is the presentative hinnẹ, often followed by a participle marking the future.

VOCABULARY

The vocabulary of biblical Hebrew is, in accordance with the limited size of the Bible, restricted, exhibiting many words from the field of religion, morals, and emotion. Loan words include those borrowed from Akkadian, as a rule through the intermediary of Aramaic, the influence of which becomes strong in later language.

[Joshua Blau]

DISCOVERY
MAIN ELEMENTS OF THE LANGUAGE OF THE DSS
SOURCES
    Biblical Texts
    Non-Biblical Texts

LANGUAGE OF THE DDS
    Spelling
    Phonetics and Phonology
    Morphology
        PRONOUNS
            Independent Personal Pronouns
            Personal Suffix Pronouns
    The Verb
        PERSONAL SUFFIXES
        OTHER FEATURES
        PAUSAL FORMS
SYNTAX
STYLE
VOCABULARY
    Native Elements
        BIBLICAL HEBREW
        MISHNAIC HEBREW
        HEBREW OF UNKNOWN ORIGIN
    Foreign Elements
        ARAMAIC LOANS AND LOAN TRANSLATIONS
        PERSIAN LOANS
        GREEK AND LATIN LOAN TRANSLATIONS
        PROBLEMATIC ELEMENTS
THE INFLUENCE OF THE DSS UPON CHRISTIAN GREEK

DISCOVERY

The discovery of the *Dead Sea Scrolls (= DSS, 1947) provided an important missing link in the development of Hebrew (= *H): a period which spans biblical Hebrew (= BH) and mishnaic Hebrew (= MH), extending from about the middle of the second half of the first century B.C.E. to 200 C.E. Before this discovery the only extant text dating back to this period (BH to MH) were fragments of the Book of *Ben Sira found in the Cairo *Genizah. Fragments found in the excavations at *Masada, however, indicate that the language used in the Genizah fragments is corrupt and does not faithfully represent the original text.

During the last centuries B.C.E., BH ceased to be a spoken language. Insofar as H was spoken, it was apparently, more or less, of the type that later emerged as MH. The literary language, which is represented by the DSS language, tried to hew as closely as possible to late biblical Hebrew (= LBH), as represented, for example, by the Books of Chronicles, which originated during the first centuries of the Second Temple period. Thus the language of the DSS should be considered as the last offshoot of LBH.

MAIN ELEMENTS OF THE LANGUAGE OF THE DSS

The DSS language, which apparently served only as a literary vehicle, is composed of the following elements: (1) BH; (2) Official Aramaic; since Aramaic (= A) had become, even before the destruction of the First Temple (586 B.C.E.), the lingua franca of the Near East and apparently also the vernacular of (nearly?) the whole of Syria-Palestine, it influenced the development of H; (3) vernacular H (the later MH) which increasingly infiltrated literary H.

The above three elements were the most important in the shaping of the language of the DSS which, however, also reveals novel traits, especially in spelling and phonology. These elements, apparently reflecting the linguistic situation in Palestine during the period in question, stem from the fact that (4) the educated classes apparently spoke Greek; (5) the vernacular of the common people was (apparently) an A dialect (or dialects) slightly different from literary A, Official Aramaic, the so-called A of the empire; (6) neither BH as transmitted, nor vernacular MH, was uniform (there being mainly two traditions: the Tiberian and the Babylonian, each with its own vocalization; sometimes the Babylonian forms emerge in the DSS); (7) there are many elements of Samaritan Hebrew (= SH) in the language of the DSS; (8) the possibility that archaic forms survived in the language of the DSS that had disappeared from H (known from the vocalization of the Jews and the reading traditions of the Samaritans) cannot be excluded.

Negative factors were also decisive in the formation of the DSS language: (9) an ever-decreasing knowledge of BH, resulting in a situation where archaic and rare words and forms of BH became obscure to the average literate Jews. In his writing, which in intention was to be BH, the literate Jew was inclined to replace obsolete words and forms with common and familiar ones. Thus a kind of basic BH, which included the above elements, came into being; (10) certain biblical words in the vocabulary of the DSS, whose meaning had become obscure (known today through modern research), not used in their original meaning but according to the interpretation given to the words by the members of the sect and their contemporaries; (11) the scrolls contain words that might have been taken over from other languages or dialects of Palestine or the neighboring territories, but have since disappeared entirely (e.g., the language of the Edomites, living in southern Judea); (12) it is possible that some new words in the DSS were common in the H or A vernacular(s) and by mere chance are not in the transmitted H and A sources.

The elements that compose the language of the DSS might have varied with the different writers. There are, for example, sources in which the role played by MH is much more prominent (e.g., the *Copper Scroll) than in other sources. The complexity of the picture that emerges from the DSS is the reason why there is as yet no solution for many of the linguistic problems in the DSS, and the outline of this language, given below, can only be tentative.

SOURCES

Biblical Texts

The complete Isaiah (1QIsaa) is one of the most important scrolls of the DSS. The language, which is "vulgar" (i.e., the intention was not to render the text exactly as transmitted), is a "modernization" of the original Isaiah, as represented by the Isaiah type of the Masoretic Text (= MT) whose language was modified so that the contemporary reader might more easily understand the text. As has been noted above, the average reader was scarcely able to understand the MT properly, and often unable to read it correctly. Therefore, copyists often substituted the contemporaneous forms for the original ones even in the case of proper nouns. For example, the form ישעיה ,עוזיה, representing the type that became common mainly after 586 B.C.E. (the destruction of the First Temple), is used instead of the original עוזיהו ,ישעיהו which represents the dominant type during the previous period.

Most of the biblical texts of the DSS are practically identical with those of the masoretic Bible. Fragments of "vulgar" texts from other books of the Bible, such as Exodus and Psalms, were, however, also found. The linguistic differences between these texts and the masoretic Bible are more or less to be attributed to the tendency to "modernize."

Non-Biblical Texts

The most extensive non-biblical texts are the Manual of *Discipline; the *Thanksgiving Psalms; the *Pesher Habakkuk, a commentary on the Book of Habakkuk; and the *War Scroll. These texts originated with a certain sect (generally identified with the *Essenes). A few fragments of non-canonical writings were also found, such as, The Book of *Jubilees and Ben Sira whose Hebrew version had until that time been practically unknown (see above Discovery). The Zadokite documents, as represented by the fragments found in the Genizah, do not reflect the original language of the DSS. (As with the Book of Ben Sira found in the Genizah, their language was also changed by the copyists of the Middle Ages.)

LANGUAGE OF THE DDS

Isaiah is the only DSS text which has been extensively dealt with from the linguistic point of view. The following survey, therefore, will be based mainly on the language of this scroll.

Spelling

The DSS employ the scriptio plena to make reading easier (there existed as yet no vowel signs) and to eliminate, as far as possible, A pronunciation. For example, לא ("not") was spelled לוא, otherwise the reader might have read it as לָא () as in A (indeed, the Samaritans in their Bible substitute the A form for the H). Plene spelling with ו abound to indicate the phones (u, o), not only where these vowels represent an original dipthong (aw) that turned into an (ō), e.g., יום ("day"; a spelling common also in the MT), or an original long vowel, e.g., שלוש ("three"; which is fairly common in MT), but also for originally short, later lengthened, vowels, e.g., עול ("yoke"; this type of spelling is rare in MT), and short vowels that were not lengthened, e.g., אוזנים ("ears"; very rare in MT), and even half vowels, e.g., חולי ("illness": = חְָלִי; extremely rare in MT). To a lesser extent, the same applies to the use of י, e.g., מית ("dead"; = מֵת in MT), אַבְנֵטְךָ=) אבניטך, "your girdle"). א was also used as a vowel letter to indicate the vowel (a), e.g., יאתום ("orphan"; the spelling is extremely rare in MT). As a word final י might be used instead of ה indicating an (e) type vowel, e.g., יעני ("he will answer"). Generally there is no real consistency in the spelling. The word "head," for example, has different spellings: ראש ,ראוש ,רואש ,רוש. The DSS share these types of plene spellings with MH, especially as preserved in the manuscripts.

The DSS, mainly, in the Isaiah Scroll and only sporadically elsewhere, developed another type of plene spelling where a digraph (two letters) indicates one vowel (like the English ea ("beat") as against i ("bit")). This type of spelling is exemplified at the end of words: וא indicated (ō), יא indicating (i), e.g., in the above mentioned לוא ("not," the א is original and the ו is added); בוא ("in it," the ו is original and the א is added); כיא ("because," the א is added). These spellings appear also in the middle of words, e.g., ראוש ,רואש ("head"), but in this case the א is practically always original, while the ו (and י) are added. As to consonant spelling, שׂ at that period turned into ס, thus the spelling שאי=) סאי "lift up") is found. There is, however, also the inverse where through a hypercorrection מאס ("to despise, reject") is spelled מאש.

Phonetics and Phonology

An outstanding feature of the language of the DSS is that the laryngeals א ,ה and the pharyngeals ח ,ע, which became weakened, are sometimes dropped and sometimes confused with each other. For instance, תאנתו =) תנתו "his fig"), יעבור =) יבור"he will pass"), מנהל =) מנחל "leader"), סלע =) סלה "rock"). These pronunciations are a characteristic of SH and Samaritan Aramaic (= SA). According to both Talmuds, the Jews of Beth-Shean and Haifa, probably influenced by the Greek vernacular, could not pronounce these phones properly. (The same applies to European immigrants in Israel, since European languages also lack these phones. They do, however, exist in Arabic and the Yemenites, therefore, pronounce them properly.) It is then possible that the weakening of these phones in the DSS occurred under the impact of Greek.

Morphology
PRONOUNS
Independent Personal Pronouns

Instead of את ("you" fem.), אתי is sometimes found. This form (rare in masoretic BH), which at first sight seems to be archaic, is probably an Aramaism (the same happened in SH). Instead of הוא ("he") and היא ("she") very often הואה and היאה are employed in the DSS. These spellings might reflect archaic forms that disappeared from masoretic BH (see above, Main Elements of… DSS, 8), but the possibility of an analogous new formation cannot be excluded. אתם = אתמה ("you" plural) is no doubt a late form parallel with the form transmitted by the Samaritans orally in reading their Bible (despite the spelling אתם pronounced אנו .(אתמה ("we") appears several times in the non-biblical scrolls (MH), in BH only once as ketib.

Personal Suffix Pronouns

The type דְבָר) ךָ) ("your word") is very often spelled plene דבריכה/דברכה (plural). The spelling disproved the theory of P. Kahle who believed that the vocalization of the Masoretic Text came into being under Arabic influence (after the seventh century C.E.!). The type ידו ("his hand") is sometimes spelled ידיו, since apparently, as in SH, יו was pronounced (o) and both ו and יו could be used indiscriminately. The type בניה ("her sons") is sometimes spelled plene בניהא/ה; the type דברכם ("your word") is often spelled דברכמה (as in SH); the type דברם ("their (masc.) word") is often spelled דברמה (by analogy); and the feminine suffix הן is sometimes spelled הנה (original? or a new formation by analogy). There is a strong tendency to use forms like רוחו =) רוחהו"his spirit"), a BH poetic form, while forms like עליו =) עלוהי"upon him") are A.

The Verb
PERSONAL SUFFIXES

Instead of the type שמרתְ ("you (fem.) watched"), the type שמרתי is sometimes used; it appears in the Song of Deborah (Judg. 5:7) whose language is archaic. The type apparently died out in BH but reentered the language under A influence, mainly in the late books of Jeremiah and Ezekiel (mostly as ketib, but not as qeri). It seems obvious, therefore, that the same influence is responsible for its emergence in the DSS (and in the Samaritan *Pentateuch). This is not the only known case where an archaic form disappeared from H and reentered the language through A, thus creating the impression of an archaic survival. As to the type שמרתמה (MT שמרתם "you (plur.) watched") it also parallels the Samaritan oral transmission (but spelled שמרתם) and is, as אתמהאתם < (see Morphology, Independent Personal Pronouns), a later development.

OTHER FEATURES

Instead of מסיר ("is taking away") (hif ʿil) there is מהסיר, which again reflects the impact of A. Spellings like ישפַל =) ישפול, "he will be degraded") indicate that an original (a) imperfect might turn into an (o) imperfect; such spellings are prevalent in MH and A dialects.

PAUSAL FORMS

Instead of the imperative form דִּרְשׁוּ ("seek"), there is דרושו, which is either an A form, or an H pausal, occurring also in BH according to the Babylonian vocalization. There seems to be little doubt that the forms of the type תזכורי ("you will remember") and אתמוכה (see below; pausal forms in the context) penetrated the DSS from MH. These forms are still found in the manuscripts (see Mishnaic Hebrew Language) and even entered the Christian Palestinian Aramaic (= ChPA) dialect from MH. The long imperfect forms of the type אתמוכה ("I would like to uphold") (expressing wish, etc.), which often take the place of the normal imperfect אתמוך, continue a trait fairly common in LBH (e.g., Ezra). The assumption of R. Meyer that the DSS employed a tense not found in masoretic BH, the so-called present future (as in Akkadian), is unfounded. However there is as yet no clear-cut solution to forms like יִשְפְּטֵנִי =) ישופטני).

Among the new noun patterns one especially worth mentioning is quṭl (= qoṭel in masoretic H) which sometimes appears as qoṭol, e.g., אוהול ("tent") = אוֹהֶל; cf., masoretic מֹלֶךְ= "Moloch" in the Septuagint; kindred forms also appear in ChPA.

SYNTAX

Although very little research has been done in the field of syntax, a few characteristics can be mentioned. Biblical syntax is employed, including the use of the conversive ו(waw), yet the copyist of Isaiah (see above Sources) occasionally substitutes a form belonging to the contemporaneous spoken idiom, e.g., instead of ואת כל אלה ידי עשתה ויהיו כל אלה ("All these (things) my hand has made") he used… והיו כל אלה ("and so all these things came to be (mine).") (Isa. 66:2). Biblical syntax requires here the imperfect plus the conversive ו. In MH the perfect is followed by the perfect plus ו; it is then the MH construction substituted for the BH one. Asyndetic relative clauses (= without the relative pronoun אשר) are still found in BH. Since they disappeared from MH and A, the writer is inclined to add the "missing" relative pronoun אשר and instead of בדרך תלך ("the way you should go") (Isa. 48:17) he creates the normal clause: בדרך אשר תלך. Sometimes he employs other means to evade the problem: in Isaiah 62:1 וישועתה כלפיד יבער ("and her salvation is as a burning torch") where there is no אשר after כלפיד he turned יבער into לפיד ;תבער thus is no longer the subject of יבער but ישועתה becomes the subject of תבער. The translation now is "and her salvation will burn like a torch."

אשר לוא plus the imperfect seems to be employed as a prohibitive in the non-biblical scrolls, e.g., אשר לוא ילך איש ("no one shall walk") which might have its parallel in MH and in the H and A letters of Bar Kokhba. The infinitive plus ל is sometimes used as a command, e.g., לשלח הואה מאתם ("They shall banish him") (LBH) and mainly negated by לוא as a prohibition, e.g., ולוא לסור ("and not to turn aside") also with אין, e.g., ואין לצעוד ("not to walk"). The same use is found in A inscriptions in Jerusalem: לא למפתח (lit., "not to open"), practically unknown in BH; it is also found in Punic (a Canaanite dialect of the Northern African coast); and it perhaps has its parallel in a certain Greek usage (found also in a Greek inscription in Jerusalem). It is impossible to pinpoint the origin of this use.

In the Book of Chronicles (LBH) the use of the accusative particle את with the pronominal suffix is generally avoided as in the Manual of Discipline (see above Sources). This is also the case in MH (as represented by the language of the tannaim only) which in this respect is a direct offshoot of the DSS.

Note בבית =) אבית "in the house"), and kindred forms, as in MH and Punic. Types like לאין שרית ("without a remnant") is to be found in LBH.

STYLE

In this area, too, more research is required. One point certainly deserves to be mentioned. The non-biblical scrolls are full of either biblical quotations, most of which are slightly different from the original, or of biblical allusions where the meaning is often not quite clear but the reference of the allusion is known, for instance, כי בשררות לבי אלך למען ספות הרוה את הצמאה "I follow my own willful heart – to the utter ruin of moist and dry alike" (i.e., everything) (Deut. 29:18) is alluded to by וילך בדרכי הרויה למען ספות את הצמאה out of which sense can hardly be made.

VOCABULARY

The vocabulary of the DSS consists of native and foreign elements (loans and loan translations).

Native Elements

These comprise BH, MH, and H of undefinable origin that does not occur in any other H source, but might be original (or a loan? see above Main Elements of… DSS, 8 and 12).

BIBLICAL HEBREW

BH. has to be subdivided according to its sources: (1) archaic BH (= ABH); (2) standard BH (= SBH); (3) LBH. BH should also be divided into: (a) words that survived unchanged; (b) words that are used in a morphologically changed form; (c) words whose meaning changed; (d) words whose meaning changed owing to a certain interpretation of their original meaning which had been forgotten; (e) words which are new morphologically and semantically, but which arise "legitimately" from BH.

ABH. אדיר ("mighty"), האזן ("to listen"), מעון ("dwelling place").

SBH. SBH is very much in evidence. אדון ("lord"), חלק ("to divide"), מערכה ("line of battle"), עת ("time"). Thanks to the DSS, MH, and SA it has been shown that the BH מסר means "to count" (> "to hand over").

LBH. Since the Hebrew of the DSS is the last offshoot of LBH (above Discovery), the presence of LBH words is not surprising, e.g., זוע ("to move") (intransitive), מדע ("knowledge?" "opinion?"), מדרש ("interpretation, study"), עמד ("to get up") instead of פשר ,קום ("explanation"), שר ("prince" (= "angel")), כוהן הרואש ("high priest"). The case of קץ ("time, epoch") is striking. Its proper meaning in LBH, "time" (and not only "end"), was (re)discovered mainly because of its usage in the DSS. These instances represent words that have survived unchanged (type a.)

BH. מָשְׁזָר (participle of hof ʿal) (= "twisted thread") appears as משוזר (obviously a participle of puʿʿal); the root נדב is used as nif ʿal (with the meaning of hitpaʿʿẹl = "to volunteer"); לַהַב ("flame, blade (of sword)") is found as לוהב. The plural of איש ("man") sometimes appears as אנושים (as if it were the plural of אנוש "man," which does not occur in BH). These instances represent words that are used in a morphological changed form (type b).

גורל ("lot") is also used for "group" and עצה ("counsel") as "council." מבקר ("inspector") goes back to the BH root בקר ("to visit" > inspect). דשא (מועד) "spring (time)" (like Akkadian and Epigraphic South Arabic) in BH is "herbage"; זרק ("javelin, dart"). These instances represent words whose meaning changed (type c).

חרישית (= "stormy wind") apparently goes back to the interpretation of Jonah 4:8, but the meaning is obviously not the original (an instance of type d).

MISHNAIC HEBREW

גודל ("thumb") = אֲ) גוּדָּל) in MH; כנסת ("assembly"); תלמוד ("learning"); מועט ("little"); מלאה ("pregnant woman" – this word occurs in the Temple Scroll); גבל ("to knead"); זעטוט ("youth (young man)"); ממון ("wealth"); נחשול ("wave"); and הלכה ("rule"?) might be MA or A. But בית משפט ("court") is perhaps a loan translation of the MH בית דין. Several technical terms of the sect also are found in mishnaic sources, e.g., קרב ("admission," lit., "to bring near"), or רבים (= "the many") which seems to be one of the names of the sect. However, רוב apparently means "many" (only?) as in BH, but not "majority" as in MH.

HEBREW OF UNKNOWN ORIGIN

א)בדן) (a kind of brocade?) whose root does not occur in BH. The roots מזז and תחם are as yet unexplained.

Foreign Elements

These are (1) A loans and loan translations; (2) Persian loans; (3) Greek and Latin loan translations.

ARAMAIC LOANS AND LOAN TRANSLATIONS

(MH vocabulary itself derives from A) דוכי ("purification") = the H root זכה. Typical for the language of the DSS is the root סרך used as a verb ("to draw up in battle order") and as a noun (meaning "order, battle order, ordinance, prescription"). It seems to be a loan word from A, but the meanings mentioned above are nearly unknown in A. The meaning of the root סדר (verb and noun), employed as a military term, is close to סרך and is apparently also A; אוחזי אבות ("intercessors") is a loan translation from A, going back to Akkadian.

PERSIAN LOANS

רז ("secret") and נחשיר ("battered") should be mentioned here. The latter shows the impact of the life at the court of the Persian governor.

GREEK AND LATIN LOAN TRANSLATIONS

Since there are no Greek and Latin loans in the DSS, it seems to be dangerous to hazard any suggestion concerning loan translations. However, if מגדל ("tower, turret") denotes a military structure and if the same holds true for the Greek πύργος and the Latin turris, there is reason to believe that some kind of connection exists between Indo-European words and Hebrew words. Even if it is assumed that the term יחד ("community") goes back to BH, the fact that the sect chose this term might have been influenced by the Greek κοιυωυία. But כנפים ("wings") as a military term cannot be taken as a sign of Latin influence.

PROBLEMATIC ELEMENTS

Several words, among them תעודה, are not entirely clear, both with regard to their meaning and with regard to their development.

THE INFLUENCE OF THE DSS UPON CHRISTIAN GREEK

Scholars found a number of terms in the DSS which parallel Greek terms in the New Testament, e.g., "sons of light" (Luke 16:8). There is reason to believe that the Greek ὲπίσκοπος, a technical term of early Christianity (> "bishop"), reflects the term מבקר ("overseer") of the sect. The Greek τάγμα found in *Josephus, designating the sect of the Essenes, seems to be a loan translation of the term סרך which, as in the compound word היחד) סרך), was employed by the sectarians as the name of their sect.

Words and phrases quoted can be traced with the help of concordances and E.Y. Kutscher's indexes.

[Eduard Yecheskel Kutscher]

INTRODUCTION
TYPES OF MISHNAIC HEBREW
GEOGRAPHICAL PROVENANCE OF MISHNAIC HEBREW
THE PROBLEM OF MISHNAIC HEBREW
THE PROBLEM OF THE SOURCES OF MISHNAIC HEBREW
    Spelling
    Phonetics
        CONSONANTS
        VOWELS
        ASSIMILATION AND DISSIMILATION
    Morphology
        PRONOUNS
            Independent Personal Prounouns
            Possessive Suffixes
            Independent Possessive Pronouns
            Demonstrative Pronouns
                Near Deictic Pronouns
                Far Deictic Pronouns
        VERB
            Verbal Roots
            Conjugations
            Prefixes and Suffixes of the Tenses
            Participle (Imperative and Infinitive)
            Verb Classes
                Strong Verb
                Weak Verb
                Geminate Verb
            Tenses
        NOUN
            Plural
        PARTICLES
        SYNTAX
        VOCABULARY
            Hebrew
            Loanwords
                Persian
                Akkadian
                Greek
                Latin
                Other Languages
                Influence of Aramaic
                    Orthography
                    Phonetics and Phonology
                    Inflection
                    Tenses and Syntax
                    Vocabulary
DIALECTS OF MISHNAIC HEBREW
MISHNAIC HEBREW OF THE PALESTINIAN AMORAIM
MISHNAIC HEBREW OF THE BABYLONIAN AMORAIM

INTRODUCTION

The destruction of the Second Temple probably brought the continuous development of biblical Hebrew (= BH) (together with its last branch, the Hebrew of the Dead Sea Scrolls) to an end. With the destruction of the religious and spiritual center, the standard literary language disappeared, and its place was taken by the vernacular, namely mishnaic Hebrew (= MH). The recent discovery of the Bar Kokhba letters, some of which are in MH, supports this view. It is, however, most likely that MH had already existed previously for hundreds of years as a vernacular. Its influence can be detected in the later books of the Bible, e.g., the Chronicles and Esther, but it was not employed as a literary language until after the destruction of the Second Temple.

TYPES OF MISHNAIC HEBREW

Two main types of MH should be distinguished: (1) The language of the tannaim, i.e., the Hebrew (= H) of the Mishnah, the Tosefta, the halakhic Midrashim, and the baraitot in the two Talmuds. (It seems, however, that the baraitot of the Babylonian Talmud were influenced by the language of the amoraim, see (2)). It may be assumed that these literary works go back to a time when MH was still spoken, most probably until the end of the second century C.E. (see below). The language of the tannaim is known (a) in the form as used in Palestine (often vocalized with the Tiberian vocalization); (b) in the form it was transmitted in Babylonia, sometimes vocalized with the Babylonian vocalization. (2) The language of the amoraim. A distinction, however, must also be made between (a) the language of the Palestinian amoraim (the Hebrew in the Palestinian Talmud and the aggadic Midrashim); (b) the language of the Babylonian amoraim (the Hebrew in the Babylonian Talmud). Since at this period (third–fifth centuries C.E.), MH was probably no longer a spoken language in Palestine – certainly not in Babylonia – it may be assumed that, as in modern H, this dialect was mixed with BH, as well as with Aramaic (= A) of the respective areas (more than tannaitic H). As a result, the H of the amoraim cannot be employed as a trustworthy basis for the study of MH (on further difficulties, see infra second drawback of Segal – The Problem of the Sources of MH).

Besides the above three categories, mention should be made of the language of prayer and benediction which also in the language of the tannaim contains elements from BH. Even in general prose the BH elements in tannaitic sources might in a few cases be quotations or allusions from the Bible rather than living elements.

GEOGRAPHICAL PROVENANCE OF MISHNAIC HEBREW

It may be assumed that MH was the vernacular only in Judea which was resettled by the Babylonian exiles in the sixth and fifth centuries B.C.E. In the rest of Palestine, especially in Galilee which had been conquered by the Maccabees (second century B.C.E.), a was apparently the only vernacular. The few A words in the New Testament also point to this conclusion, since the major New Testament figures came from Galilee. After the Bar Kokhba revolt (132–135 C.E.), however, when the Romans had nearly annihilated the whole population or sold them into slavery, the number of settlements in Judea was greatly diminished. The rabbis and their disciples moved to Galilee bringing with them their language and the tannaitic literature written in it, i.e., MH. On the other hand, their children, born in an Aramaic-speaking environment, did not continue to speak H. As R. Meir (a contemporary of R. Judah ha-Nasi) states: (שמע שחרית ערבית ומדבר בלשון הקודש הריהו בן העולם הבא כל הדר בארץ ישראל וקורא קרית (קריאת. ("Anyone who dwells in Ereẓ Israel, recites the Šemaʿ morning and evening, and speaks in the 'Holy Tongue' is assured a place in the world to come," Sif. Deut. 333 and parallels). While there were still Jews who spoke MH, its position was already shaky and was in need of some kind of strengthening. The statement of R. Judah ha-Nasi: או לשון יונית בארץ ישראל לשון סורסי למה? או לשון הקודש ("In Ereẓ Israel why Syriac (i.e., Aramaic)? Either the 'Holy Tongue' or Greek," BB 82a) shows that the language of his contemporaries was mainly A.

The few Jews who continued to live in Judea possibly still spoke H. An indication of this may perhaps be found in the statement of R. Jonathan (fourth century C.E.) from Eleutheropolis, southern Palestine, who recommended עברי לדיבור("Hebrew as the vernacular," TJ, Meg. 71b, bot.). This indicates that MH had not completely died out in this area, but in Galilee it was nonexistent. R. Johanan (the first Palestinian amora who was still a disciple of R. Judah ha-Nasi) had to emphasize that in MH the correct plural of רָחֵל ("ewe") is in a certain case (Epstein) רְחֵלוֹת and not רְחֵלִים as in BH. (His maxim was לשון תורה לעצמה ולשון חכמים לעצמן ("The language of the Torah is a language by itself and the language of the sages is a language by itself" (Hul. 137b)). The assumption that MH died out because the tannaim moved to Galilee explains why the disciples of R. Judah ha-Nasi had to ask his maidservant the meaning of such H words as מַטְאֲטֵא ("broom") (occurring in the Bible) and חֲלַגְלוֹגוֹת ("purslane") which were unclear to them (Meg. 11a). It may be assumed that the (old?) maidservant had moved from Judea to Galilee with R. Judah's household, and, therefore, spoke H. On the other hand, the (young?) disciples, who may have been born in Galilee, did not know the meaning of these words.

THE PROBLEM OF MISHNAIC HEBREW

The religious reformer A. Geiger, who was the first to write a scientific grammar of MH, thought that MH had never been a spoken language, but had been artificially created by the rabbis to facilitate their halakhic discussions. He was not the first to hold the opinion that MH was not a "normal" H dialect; some medieval Jewish scholars considered it to be a "corrupt" BH to a large extent. Since the concept of linguistic development was unknown in the Middle Ages, medieval scholars could see the reason for the differences between BH and MH only as deliberately wrought changes. Geiger, however, lived at a time when the historical study of languages and their development was taken for granted. H. Graetz, S.D. Luzzatto, and J. Levy, contemporaries of Geiger, strongly opposed his views. However, they, like Geiger, did not substantiate their arguments with tangible proofs and Geiger's view came to be accepted by all contemporary non-Jewish and some Jewish scholars until Segal refuted it convincingly.

In an article published at the beginning of the 20th century (JQR 1908), M.H. Segal showed Geiger's views to be unfounded. He demonstrated that MH was a natural outgrowth of BH (by BH is meant, besides the archaic poetic H and the standard prose, also late biblical Hebrew (= LBH) such as the language of the Books of Chronicles and the Book of Esther) and the natural link coming after LBH. As an example, consider the independent first person singular pronouns אָנֹכִי – אֲנִי (= I) both of which are found in BH. In LBH there is a distinct trend toward the use of אֲנִי. Moreover in the Books of Chronicles, which parallel the Books of Samuel and Kings to a great extent, אָנֹכִי is replaced by אֲנִי (e.g., I Chron. 21:10 = II Sam. 24:12). In MH only אֲנִי survived. Were MH an artificial language, it would be impossible to understand how the rabbis, not being modern linguists, were able to choose only the elements which belong to LBH. The situation is understandable, however, if it is assumed that MH was the natural continuation of LBH.

MH also has forms which are to be found neither in BH nor in A. Were Geiger correct in assuming that MH was an artificial creation, representing a mixture of BH and A, these novel forms in MH could not be explained, for example, where did MH get the pronoun אָנוּ ("we," found once in the Bible (Jer. 42:4) as ketib)? Clearly Geiger's opinion is in this form totally unfounded (see following par.).

The recent discoveries in the Judean Desert, especially the letters of Bar Kokhba and his contemporaries, some of which are written in MH, have dispelled all doubts as to Segal's conclusions. These letters show – as was rightly pointed out by Milik – that MH was a living natural language. As a matter of fact, however, both Segal and Geiger were right. MH was a living language in Palestine only until about 200 C.E., the time of the tannaim, but a dead language during the time of the amoraim.

THE PROBLEM OF THE SOURCES OF MISHNAIC HEBREW

Segal committed two methodical errors in his study which he repeated in the grammars of MH composed later: (1) he tried to minimize the extent of the influence of A on MH; (2) he based his work on the printed texts of MH rather than on manuscripts, which was an especially grave scholarly misjudgment.

The studies over the past decades of J.N. Epstein, H. Yalon, and S. Lieberman have shown that the printed texts are unreliable. This does not refer only to normal scribal errors, but it can be shown that during the Middle Ages the copyists, and later the printers, tried to harmonize MH with BH because they considered departures from BH in MH as mistakes. This "correcting" tendency led to a complete distortion of the linguistic structure of MH.

The following examples will prove this point. A glance at any dictionary of MH will show that the word "man" occurs in the BH form אָדָם. Since Segal's works appeared, however, hundreds of examples of the spelling אָדָן have been discovered in manuscripts of the Mishnah, Tosefta, the Palestinian Talmud, and the aggadic Midrashim (Epstein). It was corrected out of existence in printed versions, and in manuscripts where the form אָדָן does appear the beginnings of correction can already be observed (see, e.g., Ms. Kaufmann to Ber. 1:8). This phenomenon may be taken as clear proof of the widespread tampering with the printed text: the form אָדָן has completely disappeared from the printed texts on which the existing dictionaries of MH are based.

The following is another example: Segal states in his grammar that the second person singular masculine possessive (and objective) pronoun in MH is identical with the biblical form ךָ e.g., דְּבָרְךָ ("your word"). Mainly on the basis of vocalized manuscripts of mishnaic literature, as well as the oral reading tradition especially of Yemenite Jews, H. Yalon has shown that the correct form in MH is ־ָךְ, i.e., דְּבָרָךְ. The form was still known to be MH by the disciples of the medieval grammarian *Menahem b. Jacob ibn Saruq and is preserved until this day in the prayer book of the Sephardi (and Yemenite) ritual, e.g., נַקְדְּישָׁךְ וְנָעֲרִיצָךְ ("Let us sanctify you and glorify you"). In the prayer book of the Ashkenazi ritual, however, these forms have been "corrected" by the grammarians. Only in piyyutim are traces of the form still to be found, e.g., in the piyyut for *Hoshanah Rabba: הושע נא למען) אֲמִתָּךְ (למען) בְּרִיתָךְ) ("Your truth and your covenant"). Early transcriptions on the Hexapla (third century C.E.) and in the writings of *Jerome (fourth and fifth centuries C.E.) lead to the same conclusion. They superimposed it, however, on the biblical text (Ben-Hayyim). In the Sephardi communities there were also disputes as to whether this ending should be retained or dropped because the grammarians demanded the eradication of the "error." Recently, it has been shown that the second person singular feminine possessive pronoun suffered a similar fate. In manuscripts the ending יִךְ is found; thus דְּבָרִיךְ and not דְּבָרֵךְ. Both these suffixes go back to A.

Having come to the conclusion that MH, as it appears in the printed texts, is unreliable, the problem arises: On what uncorrupted source can a description of MH be based? It can also be shown that even manuscripts of the Mishnah, the Tosefta, and the halakhic Midrashim are linguistically unreliable. The problem is to find a manuscript which the copyists have changed only to a minimal extent. The same problem exists with regard to the A of the Palestinian and Babylonian Talmuds. As to the Palestinian Talmud, the problem was solved mainly by comparing its A portions to the language of the contemporary Galilean inscriptions composed in A. Those manuscripts which were linguistically the closest to the Galilean inscriptions were thus linguistically most reliable. Concerning MH, this procedure was more difficult since inscriptions or parchments written in MH, such as the Bar Kokhba letters, are quite rare.

With the aid of reliable manuscripts of the Palestinian Talmud, it is possible, however, to identify good manuscripts of MH. It may be assumed that if the A portions of the text were not corrupted by the copyists, then the H portions are also reliable. With the help of these manuscripts, the few existing H inscriptions, transcriptions of Hebrew–Aramaic words in the New Testament, in Greek inscriptions, and in the writing (transcriptions) of certain Church Fathers, it was possible to establish the most salient criteria for determining how to identify uncorrupted manuscripts. In general, the copyists harmonized the spelling conventions of MH with those of the Bible and the Babylonian Talmud. Thus, if it were possible to show that the words in a particular manuscript had spellings and forms which differed from those found in the Bible and in the Babylonian Talmud, but were parallel to forms found in inscriptions and in the Greek transcriptions from Palestine, then it would be proved that the manuscript represented Palestinian MH close to its original form.

The following are a few examples to illustrate the above methodology:

(1) In good manuscripts of MH there is the form לעזר instead of the biblical אֶלְעָזָר. This form is found in contemporary Palestinian inscriptions and in the New Testament. On the other hand, it is nonexistent in Babylonian manuscripts and sources. This shows that manuscripts with the form לעזר represent a Palestinian version.

(2) The name Shammai is always spelled שמאי in the Babylonian Talmud. In good manuscripts of the Mishnah it is spelled שַׁמַּיִ or שַׁמַּיִי. It can be demonstrated that the orthography יי ,-י- is the Palestinian representation of the final diphthong ay. (The problem of the final ḥiriq (e) remains as yet unsolved.) On the other hand the Babylonian orthography is אי-.

With the aid of several other distinguishing features, it was possible to identify several good manuscripts, in particular the following: the Kaufmann manuscript of the Mishnah (entirely vocalized), the Parma manuscript of the Mishnah (partially vocalized), the Cambridge manuscript published by W.H. Lowe (unvocalized), and fragments from the Cairo Genizah. The first two manuscripts mentioned above are vocalized with Tiberian signs, though in a vulgar manner since the punctuator, who had a "Sephardi" pronunciation, interchanged qameṣ with pattaḥ, ṣere with segol (and qameṣ qaṭon with ḥolem). The above sources represent, more or less, Palestinian tannaitic H. On the other hand, the Sifra manuscript (which is good) and certain Mishnah fragments from the Cairo Genizah, both with Babylonian vocalization, reflect tannaitic H as preserved in Babylonia.

With regard to the language of the Palestinian amoraim, the Vatican Ms. Ebr. 30 of Bereshit Rabbah, as well as the Genizah fragments of the Palestinian Talmud, were found to be reliable. Reliable sources for the H of the Babylonian amoraim have as yet to be determined.

The following description of MH is based, in the main, on the Kaufmann manuscript. Occasionally, reference will be made to Babylonian vocalized forms known mainly from the Sifra (see previous par.) and from Genizah fragments of the Mishnah (published mainly by P. Kahle and studied by E. Porath and recently by I. Yeivin).

Spelling

The spelling is more plene than that of BH. Not only the so-called long vowels (ū, ō) are spelled with ו (waw) e.g., שׁוֹמֵר("guard"), but also short and even half vowels are indicated by ו, e.g., עֳומָרִים ("sheaves") (the punctuator crossed out the ו). The same applies, more or less, to the different varieties of i-e-ε (long and short) being spelled with י (yod), e.g., לִיקְרוֹת ("to read"). Even א (ʾalep) is (rarely) used to indicate (a), e.g., שְׁיָארָה ("caravan") (also, cf., the following par.). ו and י, used as consonants, are often doubled, thus: וו, יי. The vowels ẹ, ε as word finals might be indicated by י, cf., יווני = Yavne (see the following par.). Sometimes even spellings reminiscent of the Dead Sea Scrolls are found, like לִיקְרֹאות ("to read"); the etymological א plus the ו indicating (o). As the abovementioned שְׁיָארָה indicates, א could be used as a vowel letter for a.

Phonetics
CONSONANTS

The consonantal inventory of MH is identical with that of BH. Though, undoubtedly, some change took place in their realization (= pronunciation) during the period under discussion, there is no foundation whatsoever for Kahle's assumption that the laryngeals and pharyngeals were completely lost. Nevertheless, some interchanges of these phonemes are found. It is known that in Tivon, Haifa, Beth-Shean, and in the academy of Eliezer b. Jacob ע (ʿayin) and א (ʾalep) were interchanged. According to the Babylonian Talmud, the Galileans were unable to distinguish between א (ʾalep), ה (he), ח(ḥet), and ע (ʿayin) in their a vernacular, a statement which, however, seems exaggerated. The laryngeals and pharyngeals were apparently confused mainly in the large urban centers, as a result of Greek influence. MH, as transmitted, has only been slightly influenced by this confusion and there are only a few places in the Mishnah where the amoraim are in doubt as to whether the correct reading is with א or ע, e.g., אֵיד or דיע ("festival") (Mishnah Av. Zar. 1:1).

It is quite possible, however, that the linguistic change ע < ח (ḥet > ʿayin) took place (as in Galilean Aramaic), e.g., עָג עוּגָה ("he made a circle") (Mishnah Ta'an.3:8). Final mem in non-declined words very often turns into nun אָדָן < אָדָם or הֵם > הֵן (see above the Problem of the Sources of MH). בֿ(bet = ב without dageš) and ו (waw) merged. Thus they were interchanged in manuscripts, e.g., יַבְנֶה (place name) is spelled יווני. Interchanges between ק (qop) and כּ (kap) are very infrequent. More common is the interchange בּ (bet) and פּ (pe), e.g., שְׁעָרִים = לְהַפְקִיעַ שְׁעָרִים (= לְהַבְקִיע) לְהַבְקִיַע ("to raise prices arbitrarily") (Ta'an. 2:9). Initial א (when followed by a half vowel?) is sometimes dropped (+ its vowel) cf. above לעזר אלעזר (The Problem of the Sources).

VOWELS

The vowels of MH at first glance also seem identical with those of BH. There is, however, reason to assume that some change took place, thus instead of ḥiriq qaton a type of ε (segol) was pronounced, and instead of qibbuẓ a type of ο (qameṣ qaṭonḥolem). However, even in manuscripts, very few examples of this pronunciation have survived, apparently as a result of the "corrections" of copyists under the influence of BH, e.g., חוֹצְפָּה ,הֶלֵּיל etc. = הִלֵּל (proper noun), חוּצְפָּה("ḥuẓpa"). This type of pronunciation parallels that known from the transcriptions of the Septuagint and from vocalized texts of Galilean Aramaic.

ASSIMILATION AND DISSIMILATION

Assimilation of consonants in MH occurs more or less under the same circumstances as in BH. Vowels, as in Galilean Aramaic, preceding labials tended to be realized as o (u) e.g., מְסֻבִּין < מְסִבִּין (in the Haggadah of Passover "reclining") (Ben-Ḥayyim). ר (reš) seems to have had the same effect on vowels as labials. This accounts for forms like קַרְדֹּם > ‡ קוֹרְדּוֹם (BH) ("spade"), etc. (also cf. the Greek name of the river יַרְדֵּן = Yordan (ēs)). A long ī apparently could turn a preceding half vowel (šewa (:)) into an i, e.g., בִּיסִיד (instead of בְּסִיד) ("with lime").

Dissimilation of a consonant occurs in the word מַרְגָּלִית μαργαρίτις ("pearl") and of a vowel in the Greek word נִימוֹס (from Greek νόμος ("law") (on the pattern of תּוֹךְ – תִּיכוֹן (תּוֹכוֹן‡ > ("inside," "central").

Metathesis occurs in נמיל (("port"), לְמֵן in the Palestinian form), the Babylonian form of the Greek λιμήν.

Morphology
PRONOUNS
Independent Personal Pronouns

Comparative Table(not all the vocalizations of MH are documented) Comparative Table
(not all the vocalizations of MH are documented)

Mishnaic Hebrew Biblical Hebrew
אֲנִי אֲנִי,אָנֹכִי
אַתְּ תַאְּ ,הָּתַא אתי) אַתְּ) אַתָּה
הִיא הוּא הִיא הוּא
אָנוּ אֲנַחְנׂוּ, נַחְנוּ
אַתֶּן (?),אַתֶּם אַתֶּם, אַתֶּן אַתֵּנָה, אַתֶּן אַתֶּם
הֵם,הֵן הֵם,הֵן הֵנָּה הֵם,הֵמָּה

In MH (and already in LBH) אָנֹכִי had disappeared. אַתְ as a masculine pronoun is apparently a borrowing from A. אָנוּ is an internal H development. The vocalic endings of הֵמָּה and הֵנָּה disappeared. Final ם (mem) was apt to appear as ן(nun) (see above Phonetics, consonants), therefore in both the pronoun and the verb the plural masculine and feminine forms merged (see following pars. on possessive pronouns and verb (the conjugation)).

The independent personal pronouns furnish a good example for the elements which make up MH: (1) BH; (2) A; (3) internal H development.

Possessive Suffixes

(not all the vocalizations of MH are documented)

Hebrew Language

Mishnaic Hebrew Biblical Hebrew
דְּבָרִי דְּבָרִי
דְּבָרִיךְ דְּבָרָךְ דְּבָרֵךְ דְּבָרְךָ
דְּבָרָהּ דְּבָרוֹ דְּבָרָהּ דְּבָרָוֹ
דְּבָרֵנוּ דְּבָרֵנוּ
דְּבַרְכֶן,-ם דְּבַרְכֶם,-ן דְּבַרְכֶן דְּבַרְכֶם
דְּבָרָם,-ן דְּבָרָם,-ן דְּבָרָם דְּבָרָם

Note: Instead of the ending ־ָם ,־ָן there occurs also an ending ים-, generally corrected to ן-, the י being crossed out (Epstein, Kutscher). It is found also as an object suffix of the perfect. The second person singular masculine (Yalon) and feminine (Kutscher) forms are the result of A influence. (On the interchange מ (mem) > נ (nun), see above Phonetics.)

Independent Possessive Pronouns

MH developed an independent possessive pronoun – שֶׁל (geminated ל), e.g., שֶׁלִּי("mine"). The distribution between this pronoun and the suffixed forms is still unclear as are the rules governing the use of the definite article in this case. The beginning of this development is to be found in the biblical form -אֲשֶׁר ל (= MH -שֶׁ), e.g., בְּמִרְכֶּבֶת הַמִּשֶׁנֶה אֲשֶׁר לוֹ ("in the chariot of his second-in-command," Gen. 41:43).

Demonstrative Pronouns
Near Deictic Pronouns

Hebrew Language

Mishnaic Hebrew Biblical
זוֹ זֶה זֹה,זוֹ) זֹאת) זֶה
אֵלּוּ אֵלֶּה

Instead of זֹאת which predominates in the Bible, זוֹ, found mainly in LBH, occurs in MH. It is possible that this word entered MH from another dialect. (If it is assumed that the form developed in MH from the BH זֹאת, it is impossible to explain the loss of the final ת (taw). The form אֵלּוּ perhaps developed under the influence of plural verbal forms, such as כָּתְבוּ, etc. It is unclear under what conditions the definite article is employed with the noun and the demonstrative pronoun.

Far Deictic Pronouns

Alongside the forms הַהִיא ,הַהוּא, etc., there are the following forms in MH: הַלָּז ,הַלָּה for the masculine and the feminine, (הָאֵילּוּ־הַלָּלוּ (הַלֵּילוּ for the plural. The particle אֵת with suffixed pronouns acts as a demonstrative pronoun (preceding the noun), e.g., אוֹתוֹ הַיּוֹם ("that day"). The reflexive pronoun is created by using עֶצֶם ("bone") (very much like the English "(my) self," e.g., קוֹנֶה אֶת עַצְמוֹ "he acquires himself (= his freedom)"); הוּא עַצְמוֹ ("he himself"). The relative pronoun is -שֶׁ, which appears both in archaic BH and in LBH.

Since -שֶׁ can scarcely go back to אֲשֶׁר of BH and besides is paralleled by the Akkadian ša, here too (see Near Deictic Pronouns) an H dialect different from BH may be assumed as its origin.

VERB
Verbal Roots

The verbal root pattern xyx, e.g., כָּרוֹךְ("to wrap") only emerges in MH while four radicals, e.g., from ע״ו roots of the type לְנַעֲנֵעַ ("to shake"), or by duplicating the last radical, e.g., עַרְבֵּב ("to mix") already appear in BH.

Conjugations

The puʿʿal has practically disappeared (the participle excepted). The perfect of hitpaʿʿel practically disappeared and the form nitpaʿal (corrupted in the printed editions to nitpaʿʿel) occurs instead (only twice in the Bible). It is apparently a blend of nitpaʿʿal and hitpaʿʿel. In the פי״ו verb an ʾettapʿʿal conjugation (borrowed from A) exists (extremely rare).

In addition to the hipʿil there is also a šapʿel conjugation (assumed to be borrowed from Akkadian through a) which is conjugated like the paʿʿel, e.g., שַׁחְרֵר ("to liberate"). Traces of the passive qal are found in the פ״נ verbs, e.g., ‡נוּטַּל ("taken"), etc., however it might be a recreation in MH as in modern Hebrew נֻשַּׁכְתִּי ("I was bitten") and not נִשַּׁכְתִּי which is identical with the piʿʿel (here an active form). This usage was extended to other verbs, e.g., נוצר(??) ("saved").

The exact meanings of the various conjugations still remain to be clarified. The following is a tentative description:

The qal is generally identical with the qal of BH, i.e., it can indicate a simple action (transitive or intransitive) and it can serve as a denominative even in a case like פָּרָה חוֹלֶבֶת (lit., "a milking cow"). There is, however, a conspicuous difference in the intransitive verbs. While in BH a form like גָּדַלְתָּ can mean both "you were great" and "you became great" (even "you are great"), in MH only the second meaning occurs, e.g. גָּדְלָה ("grew" < "she became great"); the first meaning has to be expressed by means of the auxiliary הָיָה, plus the participle or adjective, e.g., הָיָה גָּדוֹל.

Nipʿal also seems generally to be identical with BH, i.e., it can be a reflexive נִטְמַן ("he hid himself") and also נִשְׁאַל ("he asked for himself"), apparently in a reciprocal meaning נֶחְלְקוּ("they disputed"), but generally a passive, e.g., נֶאֱכַל ("it was eaten up"), and perhaps also with a new meaning to express perfectivity (inchoation), e.g., זָכוּר אֲנִי ("I am remembering") but אֲנִי נִזְכָּר ("it comes to my mind"). In BH qal זכר is employed in both meanings. Maybe נִכְנַס ("he entered") has to be explained the same way (cf., וַיֵּאָסֵף "he entered," Num. 11:30).

Piʿel, as in BH, expresses intensive action, meaning repeated action, or an action performed on many objects (Yalon) (cf., BH אׁתָם בַּתָּוֶךְ …וְאֶת הַצִּפֹּר לֹאׁ בָתָר (piʿẹl) וַיְבַתֵּר (qal) ("and cut them in two… but he did not cut up the bird," Gen. 15:10)); or when the work is performed by many actors, e.g., הָיו מְתַלְּשִׁין("they were plucking"); also as a denominative, e.g., מְעָשּׁנִין("to fumigate"); even in a privative sense מְיַבְּלִין ("to remove wens"), and as a causative מְיַלְּדִין ("to help in childbearing"). The piʿẹl also can serve in an intransitive meaning as an inchoative בִּיכֵּירוּ ("began to ripen"). A few cases of this last meaning already appear in BH, e.g., פִּתְּחָה ("has been opened," lit., "has opened"). In some cases the piʿẹl seems to have dislodged the qal without change of meaning, e.g., עִיבֵּר ("he passed") (Pes. 3:8), but whether it is a general feature of MH (Ben-Hayyim) has still to be established (cf., BH דֹּבֵר ("speaking") qal, but generally the piʿel is employed).

Hiʿpil, as in BH, serves as a causative מַשְׁחִיטִין ("cause to slaughter"), as a denominative הִגְרִיל ("he cast lots") as in the piʿel (also nipʿal and qal to a certain extent). It also serves as an inchoative הֶעֱשְׁיר ("he grew rich").

The hopʿal served as a passive of the hipʿil.

The hitpaʿʿel-nitpaʿʿal is mainly employed, as in BH, as a reflexive, e.g., נִסְתַּפַּג ("he dried himself"), also as an inchoative, e.g., נִשְׁתַּטָּה ("he went mad"), a reciprocal נִשְׁתַּתָּפוּ ("they became partners"), and very often as a passive נִתְגַּלָּה ("it became uncovered"), rare in BH. In contrast to BH where it serves as a denominative very often meaning "to pretend to," e.g., מִתְעַשֵּׁר ("he pretends to be rich"), in MH this meaning does not occur and in the hitpaʿʿel it means "to become rich" (cf. Hipʿil). The šapʿel is a causative (but conjugated as a piʿel).

Prefixes and Suffixes of the Tenses

As with personal pronouns, the masculine and feminine forms in the perfect of the verb also coalesced as a result of the phonological development of final ם (mem) > ן (nun), thus כתבתן – כתבתם. The loss of the feminine plural forms in the imperfect is the result of a different process. All the archaic forms of BH, e.g., imperfect forms with the ending n (ן) such as תִּשְׁמְרוּן ("you (plur.) will guard") disappeared from MH (in spite of the fact that some of them were identical with the parallel A forms).

Hebrew Language

Perfect
כָּתַבְתִּי
כָּתַבְתְּ
כָּתְבָה
כָּתַבְתָּה
כָּתַב
כָּתַבְנוּ
כְּתַבְתֵּן,-ם
כָּתְבוּ

(Note the full spelling of כתבתה). It should be noted that MH (as in the Dead Sea Scrolls) very often uses the pausal forms also instead of the contextual form. This is always the case in the hopʿal, e.g., הוּקְדָּשׁוּ לַמִּשְׁכָּן ("they were dedicated to the Tabernacle," Zev. 14:10).

Hebrew Language

Imperfect
אֶכְתֹּב
תִּכְתְּבִי
תִּכְתֹּב
תִּכְתֹּב
יִכְתֹּב
נִכְתֹּב
תִּכְתְּבוּ
יִכְתְּבוּ
Participle (Imperative and Infinitive)

The main changes in the participle are in the feminine singular only the ־ֶת ending is used: שׁוֹמֶרֶת ("guarding") (except for the ע״ו, and ל״א) ל״י) verbs to a certain extent), while the plural masculine employs, besides the ending םי- also ־ִין (A). In the imperative, the feminine plural is replaced by the masculine plural (cf., imperfect above). The participle can be negated by לֹא and not only by אין, while the infinitive is negated by שֶׁלֹּא ל, e.g., רַשַּׁיי =) אֵינוֹ רַשַׁיִי שֶׁלֹּא לַחְתּוֹם) ("it is not permitted not to seal").

Verb Classes
Strong Verb

In the qal perfect only the patterns קָטַל and חָשֵׁיכָה) קָטֵל "it became dark") have survived, while in the participle all three forms, attested in BH; קָטֵֹל ,קוֹטֵל, (e.g., דָּלֵק "burning") and the only case of יָכוֹל = קָטֹל (("he) can," "is able") appear. (Incidentally, the feminine and the plurals, not attested in BH Hebrew, are יְכוּלִוֹת ,יְכוּלִין ,יְכוּלָה.) In the "imperfect" there seems to be a tendency to turn (a) forms (of the intransitive verb) into (o) forms, cf., יִקרוֹשׁ ("it should congeal"). The spelling indicates an (o) imperfect; the punctuator of the manuscript, however, crossed out the ו (waw) and vocalized יִקרַשׁ (also see verbs ע״ח). In the hitpaʿʿel imperfect there appear, though rarely, also forms like תִּתְחַבַּר ("consort") (Avot 1:7).

Weak Verb

פ״א verbs: the infinite of qal is patterned after the 'imperfect'; לוֹמַר ("to say") etc. (cf., Spelling above), לוֹכַל ("to eat").

פ״ע verbs: note the form נֶעֱשָׂה ("it was done") (= נַעֲשָׂה in BH).

ע״ח verbs: in the imperfect and imperative the (a) turns (always?) into (o), e.g., יִשְׁחוֹט ("he shall slaughter") שְׁחוֹט("slaughter") (see above the strong verb).

ל״א and ל״י verbs: The ל״א verbs generally turned (as in A) into ל״י verbs: sometimes, however, the former spelling is retained, e.g., קָרִינוּ ("we have read"), but יִקְרֶא ("he shall read"), לִקְרוֹת ("to read"); in לִקְרֹאות ("to read") the original א appears, in spite of the ל״י form (see above Spelling); in the perfect the ending of the third person singular feminine is often ־ָת, e.g., הָיָת ("she was"). This ending, found also in BH (rarely) in the strong verb, is in BH considered mainly an archaic survival. Its emergence in the ל״י verb in MH cannot be attributed to a influence since it does not occur in the other verbal classes. It seems that this form entered MH from a non-biblical Hebrew dialect in which the original הָיָת ‡ had not become הָיְתָה. The ending ת (taw) is also found in the other conjugations but in the nipʿal there are, besides forms like נִיכְוַות ("she burnt herself"), such forms as נִיטְמֵת ("she became unclean") where the form of the original ל״א verb is identical with the feminine singular of the present. But the same form can also occur in an original ל״י form נִישְׁבֵּית ("she was taken prisoner"). Naturally, the biblical forms with the ending תָה also occur. In the participle qal there are two forms, e.g., קוֹנֶה ("he buys") and זָכֶה ("he takes possession," "he gains," "he obtains a privilege").

פ״י verbs: the infinitive of the qal is patterned after the imperfect, e.g., לֵירֵד ("to go down"). The same applies to פ"נ verbs: ןֵּתי ִל ("to give"); note forms like לִיטּוֹל ("to take") where the נ is assimilated (which is not the case in BH).

ע״ו verbs: qal, there are also participle forms like חוֹלוֹת(rare in BH) ("they (fem.) dance"); in the infinitive and in the "imperfect" also forms like לָדוֹן (also לָדוּן and לָדִין) are found (cf., ע״י verbs). There are in BH perfect qal forms like קָם ("he got up"), טוֹב ("he was good"), and מֵת ("he died"), paralleling similar forms in the strong verbs. From the second pattern only בּוֹשׁ survived ("he was ashamed"), as did מֵת. In the perfect of nipʿal forms like נָדוֹן and נִידּוֹן ("he (it) was judged or he (it) was discussed"), in the participle נִידּוֹן, are employed. There is also נָמוּךְ ("low, short") but in the Babylonian vocalization נָמוֹךְ. In the hipʿil there are forms like הוֹבִיר ("he left (the field) fallow") (patterned after פי״ו). In the geminated conjugations (piʿʿel, puʿʿal, hitpaʿʿẹl) the forms derived by doubling of the third radical (practically) disappeared; forms like הִתְכּוֹנֵן ("he intended" (from כּוּן)) are replaced (practically always) by the נִתְכַּוּן type, the second radical being geminated, as in the strong verb.

ע״י verbs: they disappeared almost entirely by (1) turning into דִּין ;ע״ו mostly appears (in the "imperfect") as דּוּן; (2) or by being transferred to the hipʿil (since the "imperfects" are identical); participle qal; שָׂם ("putting" via imperfect יָשִׂים) > מֵשִׂים (participle of hipʿil), only once in BH.

Geminate Verb

There is a tendency in the qal perfect and participle to employ the intransitive verbs with the transitive forms, i.e., they are patterned after the strong verb: e.g., גּוֹשֶׁשֶׁת("(ship) touches (the ground)" – in Hebrew it is intransitive), but רַבּוּ ("they multiplied") (intransitive form). In the imperfect the so-called A forms do not seem to occur (יִיצָּנוּ ("to keep it cool," Shab. 22:4) is not a clear-cut case). In the nipʿal the geminate verbs are generally treated as strong verbs, e.g., נִימְדַּד ("was measured"), נִימְדָּד ("is being measured"), תִּיקָּצֵץ("let it be cut off"). There seem to be very rare cases of forms like נִימּוֹקוּ ("they were defeated") patterned after ע״ו verbs. In the geminated conjugations piʿʿel, puʿʿal, hitpaʿʿẹl (as in the ע״ו verbs, see above) only the strong verb forms appear.

With the verbs הָיָה and חָיָה short form יִהְיֶה =) יְהֵא) is employed in the imperfect, while in the imperative the root הוה (A) is used often even in the A form הוִי (= H הוֵה) ("be" sing.); ֹווה (= H הווּ) ("be" plur.). The root חי sometimes appears in the participle qal as a geminate חיי) חי)) ("he lives"), like קַל ("he is easy"), but according to spelling חיה, obviously to be vocalized חָיֶה‡ (like זָכֶה above) but corrected in the manuscripts to חי.

Tenses

The tense system of BH underwent a radical change in MH. The following forms disappeared: the long imperfect of the type אֶשְׁמְרָה ("I will guard"); the short imperfect of the type יַעַל ("he shall go up"); the forms with the consecutive ו (waw) (וַיִּשְׁמֹר ,וְשָׁמַר) the absolute infinitive שָׁמוֹר. The infinitive construct only survived with the preposition ל, e.g., לִשְׁמוֹר("to guard"), sometimes even when the preceding verb governs the preposition מִן, e.g., אָסוּר מִלִּרְחוֹץ ("(he) is forbidden to wash"). The new system comprises: (1) the perfect (which also serves as a preterit);

(2) a practically new periphrastic form: היה ("be") (mainly used for the past but also for the future and imperative) plus the active and passive participle to indicate repeated, usual, concurrent, etc., action (rare in LBH).

The participle is employed as present and future. A new periphrastic form (mainly employed when the future needs a clear-cut indication, especially when in contrast to the present) came into being: infinitive דַּע מֵאַיִן בָּאתָה וּלְאַיִן אַתָּה הוֹלֵך + ל וְלִפְנֵי מִי אָתָּה עָתִיד לִיתֵּן דִּין וְחֶשְׁבּוֹן, עָתִיד ("Know whence thou art come, whither thou art going, and before whom thou art designed to give an account and reckoning"). Contrary to BH the imperfect does not denote future anymore: it turned into a modal form expressing wish or intention (in the first person) or command (in the third person). It is also used after an imperative, as שְׁמוֹר לִי וְאֶשְׁמוֹר לָךְ ("guard for me and I shall (will) guard for you") and as a subjunctive, after the relative pronoun – שֶׁ. The imperative survived apparently unchanged. The passive participle, mainly the qal of intransitive verbs, is employed with certain verbs as a kind of present perfect-present אֲני יָשׁוּב ("I am sitting (seated)"). מְקוּבָּל אֲנִי ("I have received") (rare in BH).

It should be noted that MH, as A, very often uses the proleptic suffix with verbs, e.g., the common expression y לרבי x אָמַר לוֹ רבי (instead of y לרבי x אמר רבי).

NOUN

The noun forms are generally the same as those in BH, though some became more widespread, especially some of the verbal nouns of the qal. About 15 different noun forms are used as verbal nouns of the qal, among them the noun pattern קְטִילָה should be especially noted, e.g., אֲנִינָה ("grief"). This noun pattern in BH as a verbal noun (e.g., אֲכִילָה ("eating")) is rare, in the Mishnah, however, there are 130 examples. Its influence was so great that it was able to change the biblical form of שְׂרֵפָה ("conflagration") to שְׂרִיפָה. In the ל״י (and ל״א) verbs this form may appear in the קִטְיָה. pattern, e.g., בִּרְיָה ("creature, creation"), קִרְיָה. ("reading"), etc. Though rare, the form קְטֵילָה is also found, such as כְּנֵיסָה ("entrance"). The form קְטָלָה is also rare (though common with verbs that denote sound), e.g., צְוָוחָה ("shouting"). A new form is גָּזֵל ("robbery"), חָנֵק("strangulation"). Verbal nouns with suffixes are also found, e.g., פִּדָּיוֹן ("redemption"). (The word does not occur in the absolute state in the Bible.) The number of A patterns is relatively small, e.g., כְּלָל ("general rule"), פְּרָט ("specification"); with the prefix מ (mem): מִכְנָס ("bringing in"), showing that A had a minor influence in this field.

The verbal noun of the piʿʿel is קִטּוּל. or קַטָּלָה (both BH but the latter is a borrowing from A). In the hipʿil also the a form הַקְטָלָה (already in the Bible) predominates along with הֶקְטֵל, e.g., הֶקְטֵר ("burning" ("of offering")). In the ל״י verbs, the form in Babylonian sources is, e.g., הוֹרָאָה ("instruction"); whereas in Palestinian sources it is הוֹרָיָה. The form הֶקְטֵל, e.g., הֶקְטֵר, is, in fact, identical with the absolute infinitive in BH. (As in BH, the segol is an allophone of pattaḥ). In the Babylonian vocalization it may appear both as הַקְטֵל and הִיקְטֵל (in certain cases). The passive and reflexive conjugations do not have their own verbal nouns and employ the verbal nouns of the corresponding active conjugations, e.g., וִידּוּי ("confession of sin") from לְהִתְוַדּוֹת. It should be noted, however, that the nipʿal infinitive הִכָּרֵת, occurring in the Mishnah also as כָּרֵת ("extermination"), serves as a verbal noun; even a plural form occurs כָּרֵתוֹת.

The form with the ending ־ָן (-ān) is a nomen agentis (the agent) noun pattern which is peculiar to MH. In Palestinian manuscripts, these appear mainly as גּוֹזְלָן ("robber"), רוֹצְחָן ("murderer"), etc. (The vocalization is not uniform.) In Babylonian sources mainly the forms גזלן and רצחן occur. The origin of this form is still unclear. The nomen agentis for qal of the קָטוֹל pattern, e.g., לָקוֹחַ ("buyer") might be of A origin. It should be noted that MH tried to develop a special form to represent the result of an action, namely קְטִילָּה (practically nonexistent in BH). The only example of this form is חֲתִיכָּה ("a piece") alongside the verbal noun חֲתִיכָּה ("cutting").

Alongside of the construct there is the paraphrastic (the circumlocuted) construct state which uses the particle של. As Yalon has demonstrated, this word was attached to the nomen rectum (if this was determined) and contained the definite article, e.g., ‡רִבּוֹנוֹ שֶׁל הָעוֹלָם = רִבּוֹנוֹ שֶׁלָּעוֹלָם ("master of the world"). How and when של was separated from the noun and ceased to contain the definite article is not entirely clear. In the Bar Kokhba letters של is separated from the following word, which, however, has the definite article. This shows that the dialect of the Bar Kokhba letters is not identical with MH as it is known today.

In the של phrase there are four types. In three of them של includes the definite article:

(a) הַיַּיִן וְהַחוֹמֶץ שֶׁלַּגּוֹיִם ("wine or the vinegar of Gentiles") (Av. Zar. 2:3). (b) with the prolectic suffix אֵידֵיהֶן שֶׁלַּגּוֹיִם "thefestivals of the Gentiles"). The difference in meaning of these two constructions is not entirely clear. In each phrase both nouns are determined. (c) שֶׁלַּזָּהָב (= נברשת) נִפְרֶשֶׁת ("a golden candlestick") (Yoma 3:10). (d) לָשׁוֹן שֶׁלִּיזְהוֹרִית ("a thread of crimson wool"). In each of these phrases both nouns are undetermined. The reason for the difference between the two last constructions is not clear.

Plural

Besides the plural with ־ִין ,־ִים, and וֹת-, a plural with the ending ־ָאוֹת in Babylonian sources, ־ָיוֹת in Palestinian sources occurs, e.g., מֶרְחֲצָיוֹת = מֶרְחֲצָאוֹת ("bathhouses"). The plural of nouns ending in וּת- is not ־ֻיּוֹת, as in the Bible, but ־ִיּוֹת, e.g., מַלְכִיּוֹת ("kingdoms"). A double plural of compound nouns, such as, רָאשֵׁי שָׁנִים ("new years") occurs (cf., for example, the form אַנְשֵׁי שֵׁמוֹת, found in Chronicles, to אַנְשֵׁי שֵׁם("famous men") which appears in Genesis).

The rules governing the use of the definite articles are still not entirely clear. It should, however, be pointed out that a noun with an accompanying adjective generally does not take the definite article, e.g., לַיְלָה הָרִאשׁוֹן ("the first night"). Other usages, such as, הַכֹּהֲנִים גְּדוֹלִים ("the high priests") which appear to be exceptions to the rule require further investigation (cf., הכהן גדול in the Dead Sea Scrolls).

PARTICLES

While there are many new adverbs and conjunctions, such as, בִּנְתַּיִם ("meanwhile"), כְּדֵי ("in order to"), עַכְשָׁיו ("now"), כֵּיצַד ("how"), noteworthy is אבית instead of בְּבֵית. It seems that the biblical prepositions have remained to a greater extent than the other particles, as in the case of the language of the Dead Sea Scrolls and Punic. Some usages which should be especially noted are the following: -ל is used to a great extent for -ב, e.g., הָיְתָה תְרוּמָה לְתוֹךְ פִּיו ("the terumah was in his mouth"). Many verbs take either one of the following prepositions: ב or ל. ל also indicates the accusative (rare in BH but common in A). The prepositions עד – על interchange (as they do in Galilean and Samaritan Aramaic). The copulative ו (waw) sometimes acts as an explicative ו (waw), e.g., מְבָרֵךְ עַל הַטּוֹבָה וּמֵעֵין עַל הָרָעָה ("A man should say the benediction for good fortune regardless of any consequent evil") (in German: und zwar).

In particles of negation besides אֵין ("not"), employed in nominal sentences including participles (note the declension: אֵינִי, etc. אֵינָהּ / אֵינוּ – "he/she is not"), also לֹא is used to negate participles. The A loan לָאו is used mainly in the phrase אִם לָאוq ("if not"). אֵי ("אֵין") occurs apparently only before א, e.g., אֵי אַתֶּם ("don't you"). The expression "yes," which is absent in BH, in MH appears as הין (from A).

SYNTAX

Owing to the radical changes that occurred in the tense system of MH (see above), the syntax of MH looks very different from that of BH. However, since research in syntax has to be based on good manuscripts (see The Problem of MH), the picture is as yet not entirely clear.

The following may more or less be stated: in the verbal sentence generally the verb seems to precede the subject but not always. A verb can take a verbal complement in three ways: (a) infinitive plus ל (as in BH); (b) the participle הִתְחִיל בּוֹכֶה ("he started weeping") (rare in BH);

(c) a relative clause, צָרִיךְ שֶׁיֹּאמַר ("he must say"). In the past conditional the construction participle plus הָיָה is preferred, אִלּוּ (negative לוּלֵא) opening the sentence, e.g., אִילּוּ הָיִיתִי יוֹדֵעַ לֹא הָיִיתִי נוֹדֵר ("Had I known (that this was so) I would not have made my vow"). Interrogative sentences which expect a negative (?) answer begin with כְּלוּם. Relative sentences are more numerous than in BH since in MH subordinate clauses are used instead of the biblical infinitive plus ב or כ (not occurring in MH).

In the comparative sentence often יוֹתֵר is added, e.g., רַע יוֹתֵר מִסודמיין ("worse than the Sodomites") (Tos. Shab. 7:23).

While in BH the passive is used almost only if the agent is unknown (with very few exceptions), in MH it seems to be employed even if the agent is known, e.g., הַנּוֹלָדִים מִן הַסּוּס ("all offspring from a horse," lit., "born by a horse"). The agent is expressed by מן and by ל, e.g., נֶאֱכָלִים לַכֹּהֲנִים ("are eaten by the priests"). As noted, the syntax of MH has to be restudied on the basis of good manuscripts.

VOCABULARY

A great part of BH vocabulary disappeared from MH including even words indicating close relation, דּוֹד ("uncle") or parts of the body, such as, בֶּטֶן ("belly"), בֹּהֶן ("thumb") was replaced by אגודל) גּוּדָּל). As is well known these two fields are the most resistant to change in every language. Less amazing is the fact that vocabulary used only in the poetic parts of BH did not generally survive in MH, e.g, חָרוּץ("gold"). The vocabulary of MH is composed of the following elements: (1) Hebrew; (2) loanwords from Persian, Akkadian, Greek, Latin, and Aramaic.

Hebrew

The Hebrew element has many facets:

(a) BH whose meaning remained the same, such as, יָד ("hand"), רֶגֶל ("foot"), בַּיִת ("house"), מַטֶּה ("staff"), יָצָא ("to go out"), שָׁמַע("to hear"), רָאָה ("to see").

(b) BH words which took on a different form (in the following examples the first word is the biblical form and the second the mishnaic): מַשּׂוּאָה – מַשְׂאֵת ("flares," "fire signs"), לְכַשֵּׁל – לְהַכְשִׁיל ("to cause to stumble"), יַמִּין – יָמִין ("right hand") (Aramaic?), זֹוֹג – זָג ("grape-peel"), חַיִל – חֵיל ("surrounding wall"). Some words found in the Bible are only in the singular whereas in MH they occur also in the plural, e.g., שְׁמִיטִּים – שְׁמִטָּהֹ ("sabbatical year"). In particular this is the case with collective nouns, such as פֵּרוֹת – פְּרִי ("fruit"), יְרָקוֹת – יָרָק ("vegetable"). Some words found only in the plural in the Bible occur in MH in the singular. such as, בָּטְנָה =) בּוֹטְנָה – בָּטְנִים "pistachio"). Verbs, such as רום > תרומה > תרם)תרם)), from the BH root רוּם, in hipʿil לְהָרִים ("to raise"), with the preformative ת (taw) formed the noun תְּרוּמָה ("heave offering"). MH derived from תְּרוּמָה a new root תרם which is now used instead of BH לְהָרִים. Nouns were formed from verbs, e.g., וַעַד ("meeting place") < הִוָּעֵד ("to meet"), וִדּוּי ("confession of sin") < הִתְוַדּוֹת("to confess") where the biblical aversion to waw as the first radical did not apply anymore.

(c) Some nouns which apparently changed their gender under a influence, e.g., כּוֹס ("goblet") which became masculine while שָׂדֶה ("field") became feminine.

(d) A biblical element which changed semantically but not morphologically. Some words are concrete in the Bible and abstract in MH, e.g., נָהוֹג "to lead," in MH "to behave." Similarly גָּזוֹר in MH means only "to decide" and not "to cut." עוֹלָם in the Bible means "eternity," but in MH "world." Some words were semantically restricted, e.g., צְדָקָה "righteousness" and "charity," in MH means only "charity." חַג ("holiday") refers only to "Sukkot" in MH and עֲצֶרֶת ("assemblage") only to "Shavuot." It is sometimes difficult to decide whether a particular biblical root changed its meaning or the root in MH is simply homophonic, for example, פָּסוֹל in the Bible means "to hew," with the derived noun פְּסוֹלֶת ("refuse") (cf., נְעוֹרֶת "tow"). From פְּסוֹלֶת a denominative qal verb was formed ("to declare unfit"); there may, however, be a different root here (as found in Arabic). But the root לְהִתְחַטֵּא ("to ingratiate himself") is certainly not identical with the Hebrew root חטא ("to sin"), but is an A root.

(e) Non-biblical Hebrew elements. It is certain that the Bible does not contain the whole vocabulary of the biblical period, as shown by personal names and inscriptions (cf., the word זדה in the Siloam inscription (as yet unexplained) and נצף appearing on weights (meaning apparently "half," cf., Arabic). Therefore at least some of those roots which cannot be proven to be foreign loans are probably survivals from the biblical period and only incidentally did not appear in the Bible. This is, of course, impossible to prove in most cases. Sometimes, it cannot be determined whether the form originated in h (or in a Hebrew-Canaanite dialect) or in a neighboring dialect (such as Edomite (?)). It seems probable that most of these words are of H (or Hebrew-Canaanite) origin. Consider, for example, the root חָזוֹר ("to return") (in Eastern Aramaic הדר – חדר; in Western Aramaic חזר, apparently a borrowing from Hebrew). This root may have reached H through one of the Canaanite dialects. On the other hand, the root of לְהִתְעַכֵּב ("to be delayed") is less certain. (There is no certain parallel in the other Semitic languages.) The certainty is greater for agricultural terms or for parts of the body, e.g., שָׂרָף) סָרָף) ("resin"), which do not occur in other Semitic languages. The form of the word טְחוֹל ("spleen") shows its Hebrew-Canaanite origin. Since the Arabic cognate is ṭiḥāl and ā appears here as ō, as in Hebrew-Canaanite (a change which did not occur in A loanwords), the H origin of the word seems more or less to be certain. The root מסוק ("to harvest olives") is probably Hebrew since it has no cognate in the other Semitic languages. It was only by chance that these roots did not occur in the Bible, or maybe they were current in a different H dialect (regarding קוסמ compare the biblical root נקוף with a similar meaning) but not in BH. In order to clarify the relationship between MH and BH, and especially A, the vocabulary of the former should be studied thoroughly on the basis of excellent manuscripts and according to different fields in semantics.

Loanwords
Persian

The Persian hegemony in Palestine lasted only 200 years and Persian consequently did not leave a strong mark. Administrative terms such as גִּזְבָּר ("treasurer"), already found in the Bible, occur, but not מַרְכוֹל (the Palestinian form) – אמרכל (the Babylonian form) ("a high official"). The word וֶרֶד ("rose") seems to be Iranian. The fact that the word begins with ו (waw) points to its non-Hebrew origin (but see above Vocabulary (b)).

Akkadian

Most Akkadian words in MH were borrowed through an A intermediary. Some words, however, do not appear in A. The Akkadian-Sumerian אַפַּר ("meadow") is hardly found in the A dialects. On the other hand אֶמָּתַי ("when"), parallel to BH מָתַי, is found in several a dialects. Many Akkadian mercantile terms, such as, שְׁטָר ("writ"), גֵּט ("writ (of divorce)"), תַּגָּר ("merchant"), אָרִיס ("tenant farmer") from the Akkadian root erēšu ("to plough") have entered MH, as have terms from the material culture, such as, דַּף ("page") (of Sumerian origin). The root זוּז ("to move") is also of Akkadian origin. It is possible that the meaning of לָקוֹחַ ("to purchase"), found mainly in MH, is an Akkadian calque (loan translation). That is apparently why when לָקוֹחַ ("to take") also acquired the new meaning "to purchase," the BH לָקַח אִשָּׁה ("to take a wife," i.e., "to marry") changed in LBH to נָשָׂא אִשָּׁה.

Greek

Many administrative, religious, mercantile, material culture (excluding agriculture), and even everyday words were borrowed from Greek. From the Greek word זוֹג ("yoke") a denominative verb was formed; similarly אָוֵיר) אֲוֵיר?) ("air") and הֶדְיוֹט ("simple person") are Greek. There are mercantile terms: פִּינְקֵס ("account book") and סִיטוֹן ("wholesale provision dealer"): household terms: קָתֶדרָה ("chair with a back"), and פְּרוֹזְדוֹד (corrupted in the printed versions פְּרוֹזְדוֹר) ("vestibule"): administrative terms: סֶנְהֶדְרִין (Greek "assembly"), and לְקַלֵּס ("to praise," mainly a king or a high official); urban terminology: מטרופולין ("city"), פלטין ("palace"), and לְמֵן (נמל in Babylonian sources of MH ("port")); food terms: כְּרוּב ("cabbage"). The expression (יָפֶה (דָּרַשְׁתָּ ("you have (well) explained") (in BH יָפֶה = beautiful) seems to be a calque, as is apparently הִשְׁלִים ("he did") (Lieberman).

Latin

The few Latin loanwords are from the administrative and military spheres, e.g., לִיבְלָר ("scribe"), לִגְיוֹן ‡ ("legion"), נוֹמְרוֹן ("troop of soldiers") סרטה – אסתרטה ("street"), קָרוֹן("wagon"), סַפְסֵל ("bench"), and טַבְלָה ("table").

Other Languages

Assuming that the language of the Edomites who settled in Palestine was closer to Arabic than to H (there is, however, no proof of this), it may be hypothesized that the word שׁוֹבָךְ ("dovecote"), Arabic šubbak ("window"), was borrowed from Edomite. The Arabic š was taken into Hebrew without the linguistic change to שׂ. According to an opinion in the Talmud, the expression יוֹנֵי הוֹרדסיוֹת (the second word appears in various forms) in the Mishnah means "doves of the king Herod," which (according to Josephus) he raised in his home. חוֹטֶם ("nose"), from a rare biblical root, brings to mind the nose ring (חֲטָם) of the camel. This word may have come into Hebrew from the language of a people that still employed camels (from Edomite?); the assumption is, of course, purely speculative.

Influence of Aramaic

Unlike the above languages whose influence on MH was felt mainly in the vocabulary, A had a far-reaching impact and left its mark on all facets of the language, namely, orthography, phonetics and phonology, morphology including inflection, syntax, and vocabulary. There is room for investigation as to whether MH was a Hebrew-Aramaic mixed language. This question may be posed owing to the fact that A had a pervading influence in all spheres of the language, including inflection, which is generally considered to be impenetrable to foreign influence. It is possible, however, that because of the symbiosis of A and Hebrew-Canaanite the two exerted a mutual influence (see especially phonology).

Orthography

All of the peculiarities mentioned above as being in MH are found, more or less, in the Palestinian Aramaic dialects as well, especially Galilean and Christian-Palestinian Aramaic, and even in the eastern dialects.

Phonetics and Phonology

The fact that the consonantal phonemes (according to biblical A also the vocalic phonemes) are from a synchronic point of view identical in both languages – a phenomenon without parallel often even in different dialects of the same language – is noteworthy. There is reason to believe that this is due to Hebrew-Canaanite: from A inscriptions it is known that there were several phonemes in A which did not exist in Hebrew-Canaanite. Common to H and A are the double realization בג"ד כפ"ת (b g d k p t); the weakening of the gutturals to a greater or lesser extent in most of the A dialects; and common assimilation and dissimilation phenomena (with regard to ר (reš), especially in Galilean Aramaic).

Inflection

The independent personal pronoun אַתְּ ("you" masc.) and the possessive pronouns ־ִיךְ ,־ָךְ (see above Pronouns) are clear indications of A influence. With regard to the verb, the influence was weaker. The A root הוה appears maybe even with an A vocalization (see above). The loss of the puʿʿal is paralleled in A, whereas the hopʿal still exists as opposed to the A dialects where it disappeared (with the exception of the early dialects). The rare occurrence of the ʾettapʿal, the development of the nitpaʿʿal, and the rejection of forms such as תִּשְׁמְרוּן (see above conjugations) point to an anti-Aramaic trend. A influence was less felt in the noun patterns.

Tenses and Syntax

The tense system completely parallels that of Galilean Aramaic and is close to that of Christian-Palestinian and Samaritan Aramaic. It is also similar to that of Eastern Aramaic. The assumption that the whole tense system is influenced by A seems to be inescapable. Note, however, that biblical Aramaic and the old A inscriptions show that this system is not original with A. Even though there still is no real comprehensive study on the syntax of MH and the Western Aramaic dialects, there seems to be a far-reaching parallelism between them.

Vocabulary

It is clear that A influence is considerable in this category. Absolute proof is provided by loanwords having an A root consonant which differs diachronically from the Hebrew cognates (ת ,ע ,ט ,ד), or by loanwords in which a difference arises because of the Hebrew Canaanite vowel shift ā > ō. Thus, for example, הִתְחַטֵּא ("to ingratiate"), אירַע ("to occur") < A ערע, Hebrew-Canaanite ערץ ‡ are all A; similarly, אֶילָּא ("but") אֶן לָא = אִם לֹא in BH, שָׁעָה ("hour"). Even in the numerals there are A elements, e.g., שְׁתוּת ("a sixth") and תּוֹמֶן("an eighth"). As is well known also the numerals are most resistant to penetration of foreign elements.

In other cases the decision may be in favor of an A influence, e.g., אֶמְצַע ("middle"), מָמוֹן (?) ("money"), and many more. There is still no up-to-date work on this subject. All the studies published in this field are unreliable.

There are also many calques, such as, סָגַר = אָחַז ("he closed"). Similarly the fact that in MH כּוֹס ("goblet") is masculine and שָׂדֶה ("field") is feminine goes back to A influence.

Due to A influence there are occasionally in MH words which are archaic in the Bible (but in general such words disappeared from MH), e.g., עוֹנָה ("time"), – יְמוֹת ("days") as in the phrase יְמוֹת הַחַמָּה ("the sunny season").

A biblical word might change in form because of A influence, e.g., גֵּיא (בֶן) הִנֹּם in the Bible, but גֵּיהִנָּם ("Gehenna") in MH (with a different meaning). This is the traditional pronunciation in several Jewish communities.

DIALECTS OF MISHNAIC HEBREW

The early state of affairs as represented by the manuscripts will be discussed here and not the differences between the living traditions of the different Jewish communities (mainly the Yemenite, Sephardi, and Ashkenazi). It is certain that there were differences between the Babylonian and Palestinian traditions. It is even possible to assume that archaic forms which later changed in the Palestinian tradition occasionally remained in the Babylonian tradition. Consider the following example: according to the transcription of the New Testament it is known that the old form of רבי ("rabbi") was רַבִּי. This form was preserved in the Babylonian vocalization tradition, but in Palestinian manuscripts the vocalization is (רֶבִּי (רִבִּי and even רְבִּי. Greek transcriptions from Palestine, and later transcriptions in Italy, prove that the first two forms are correct Palestinian forms. They were also preserved in the traditions of various communities. The form מוּעָט ("small part") is found mainly in Babylonian sources. The normal form מְמוּעָט is found mainly in Palestinian manuscripts (also מועט). However in the Dead Sea Scrolls the Babylonian form מועט occurs. A clear difference between Palestine and Babylonia is indicated by such forms as הוֹדָאָה ("thanks") (Babylonian) as opposed to הוֹדָיָה (Palestinian). Similarly גזלן (Babylonian) and גּוֹזְלָן (Palestinian).

It seems that even in Palestine there were dialectical differences and though the indications are few concerning the vocabulary, the evidence of the Talmud on certain points may be accepted. רְפָפוֹת ("shutters") were called by one tanna רְעָדוֹת. Besides this, it is difficult at the moment to find other differences, such as, the interchange ע – א (ʿayin) – (alep) attributed to the academy of Eliezer b. Jacob. The MH of the Bar Kokhba letters is slightly different from that which has been transmitted. של is not connected to the word following it. The nomen rectum, however, has the definite article (as opposed to the situation in the printed editions of MH texts). Instead of את, there is (as in Punic) ת, e.g., תכבלים = אֶת הַכְּבָלִים ("the chains") (perhaps this form will be discovered in good manuscripts). The word אזי ("then") found in these letters is not present in normal MH.

There seem to be traces of an H dialect which was not identical with BH. If this is not assumed, then it is difficult to explain the exclusive use of זו instead of זֹאת (zō + t) since there is no way of explaining the loss of the ת (taw). It is preferable to assume that זו came to predominate in MH from another H dialect in which this archaic form existed. (וז already occurs in the Bible.) The forms הָיָה ("she was") and קָנָה ("she bought") are even more to the point (see above weak verb). The regular biblical form הָיְתָה and קָנְתָה developed from קָנָת‡ + ā which was taken over from the other verbal classes. It is impossible to understand how a retrogression would occur in MH; these forms are thus better explained as intrusions from a dialect in which the process קנת > קָנְתָה did not take place. (Survivals of the archaic form occur in the Bible and in the Siloam inscription.)

MISHNAIC HEBREW OF THE PALESTINIAN AMORAIM

This dialect has been studied on the basis of Vat. Ms. Ebr. 30 of Bereshit Rabbah. On the one hand it has been found to have a considerable mixture of BH and on the other to contain independent forms that are found in MH but not in tannaitic sources. (They occur in very few cases and must have been corruptions). Thus, זאת occurs as an adjective, e.g., הלבנה הזאת("this moon"). The far deictic pronouns הַלָּה ("that") and הַלָּז ("that one") disappeared and were replaced by אוֹתוֹ ("him"), etc. These changes are to be regarded as internal H developments, though the last was perhaps influenced by A. The ending ת is sometimes found in the third person feminine perfect in verb classes other than ל״ה (A influence). In the imperfect first person singular, the first person plural form is sometimes employed (as in Galilean Aramaic). This usage is found only once in the Mishnah. As in BH the construct infinitive without -ל occurs, e.g., בּוֹא + מִן) מִבּוֹא). As opposed to MH the following differences should be noted: (1) internal H development; (2) admixture of BH; (3) increased A influence.

MISHNAIC HEBREW OF THE BABYLONIAN AMORAIM

This dialect has not yet been studied (see below). The word אשפה ("trash") as against אַשְׁפּוֹת (in the tannaitic Hebrew) may point to independent development.

[Eduard Yecheskel Kutscher]

Kutcher's description is still valid in its main features. However, since the 1970s research in MH has made considerable progress. While the description above is based on the Mishnah according to Ms. Kaufmann, in recent years other MSS of the Mishnah have been described, such as Paris, Parma 497, Deinard, Maimonides' Autograph, and Genizah fragments. Haneman's description of the verb system in Ms. Parma 138 can serve as a model for the "classic" MH verb system. Other tannaitic as well as amoraic sources were investigated, such as the Tosefta, Sifra, Palestinian and Babylonian Talmuds. Traditions contained in old sources and oral traditions were described, such as Yemen, Aleppo, Tunisia, Italy, early and late Ashkenazi traditions, the Karaite tradition, and others. The Babylonian Punctuation tradition is presented in detail in Yeivin's monumental work. Syntax is described according to Ms. Kaufmann of the Mishnah and there is a full description of the tense system.

In these descriptions many features were recognized to be typical of MH. The following are a few examples: It was proved that a doubling of ר was common in certain traditions of MH. The relative pronoun ש is vocalized with sheva in some circumstances, such as שְהוּא (these two phenomena are very rare in the Masoretic Vocalization of the Bible). Some conjugations of the verb (or modifications of old conjugations) were established: nuf ʿal as a variant of nif ʿal in verbs I-y and I-n, e.g. נֻטַּל (instead of the common נִטַּל); pŒ > el and nitpŒ > al (instead of paʿʿel and nitpaʿʿal), e.g., מְזָמֵן; nitpaʿʿal in participle can take the form נִתְפַּעֵל (instead of the common מִתְפַּעֵל). For the meaning of the conjugations, it was claimed that the hif ʿil can serve for the same meaning as the qal (as it is claimed of the piʿʿel).

In the field of vocabulary, Moreshet's lexicon lists and discusses all the verbs in MH not found in the Bible. According to his findings, there are about 500 new verbs in MH, of which two-thirds can be attributed to Aramaic influence. In this field mention should also be made of the Historical Dictionary Project of the Academy of the Hebrew Language, which produced a full concordance of tannaitic and amoraic literature according to reliable MSS (of other periods). This concordance gives an accurate list of the vocabulary of MH and serves as a fundamental tool of research in this field.

The richness and variety revealed in so many reliable sources enabled Bar-Asher to sort and arrange MH sources according to two criteria (he deals exclusively with the Mishnah, but in fact his observations are valid for all tannaitic and amoraic sources): (1) Palestinian vs. Babylonian branches, e.g., while in Palestinian sources (such as Ms. Kaufmann of the Mishnah) we find the verb נתאלמנה, in Babylonian sources (such as quotations from the Mishnah in the Babylonian Talmud) the verb is נתארמלה. Although the last verb was probably borrowed from A, it is an ancient borrowing, as it occurs already in the Hebrew of the Dead Sea Scrolls. (2) Western vs. Eastern traditions, which differ mainly in the realization of written texts; e.g. while in the Western traditions a doubling of ר is almost nonexistent, it is quite common in the Eastern traditions of MH. Many of these differences seem to go back to ancient times and may have existed when MH was still a living tongue.

[Yochanan Breuer (2nd ed.)]

INTRODUCTION
THE LANGUAGE OF PIYYUṬ
SAADIAH GAON
    Arabic Influence
SPANISH HEBREW POETRY
    Influence of Arabic on the Language of Secular Poetry
ORIGINAL PROSE WORKS AND TRANSLATIONS
    The Components of Arabic-Influenced Hebrew
        LEXICON
        NOUN MORPHOLOGY
        VERB
        SYNTAX
MAIMONIDES
ITALY
THE KARAITES
    Samaritans
PROVENCE AND NORTHERN FRANCE
ASHKENAZIC AND RABBINIC HEBREW
    The Influence of German on the Hebrew of the Jews of Ashkenaz and the Influence of Yiddish on Hebrew in Poland
    The Role of Mishnaic Hebrew in Ashkenazic Hebrew
    The Role of Biblical Hebrew
    The Role of Aramaic
    The Link With Arab-Influenced Hebrew
    Influence of Hebrew on Yiddish
    Haskalah and Medieval Hebrew

INTRODUCTION

After Hebrew as a spoken language was replaced by Aramaic, it became a written language whose history is from and for books alone. The principal sources for the writers were Biblical Hebrew and Mishnaic Hebrew and these met the needs of all forms of written expression: religious and secular poetry, letters, books on science, and philosophy.

Hebrew became a second language, existing side by side with the vernacular languages spoken by Jews wherever they happened to be. Such a duality was quite normal in the Middle Ages; spoken Arabic existed alongside classical Arabic, other languages were spoken where Latin was the literary medium. Although it became a written language, Hebrew did not remain petrified, limited to passages quoted in their original form and meaning, but lived "an active life" in written texts. New topics, whether in original writings or in translations, necessitated an expansion of the language, especially in the coining of new terms for concepts and subjects not found in the Bible, the Mishnah, or the Midrashim, e.g., philosophy, medicine, etc. Responsa which had to deal with everyday subjects, not found in earlier halakhic responsa, also led to linguistic innovation, especially in vocabulary. Since it was a written language, many new forms were invented for literary purposes: rhetorical language and stylistic embellishment, especially in poetry. It is difficult to evaluate the changes on linguistic grounds alone, particularly in poetry; the language was, as it were, raw material for stylistic variation.

In a living language which serves as a natural means of spoken communication, an innovation is any new form which carries a specific meaning (morphological-semantic innovation). Innovation of this kind occurred in the written language in books of science, especially as translations of new concepts which had previously been unknown to the Hebraic world and had no equivalents in Hebrew, e.g., agron (more correctly egron; "a dictionary"), mahut ("essence"). New meanings were added to existing words; this is a common feature of poetry as a means to enrich the language. In piyyuṭ, though not only there, use was made of the system of "alternate forms," whereby existing words could change their form – according to regular patterns of analogical formation, and also irregularly – without any change in meaning. These are morphological-stylistic changes, but not semantic. This technique is generally foreign to the spoken language where every form has its own specific meaning.

The linguistic changes of the written language, unlike those of the spoken tongue, do not take place of their own accord, through the operation of analogy, leveling, attraction, etc. They owe their existence to the needs of artistic and stylistic embellishment and are premeditated rather than spontaneous (as will be explained below). They include changes in frequency – rare words become common – sometimes because of the different frequency in the language of influence, sometimes because of a deliberate choice of words felt to beautify the language. Some innovations rose from a linguistic understanding of the processes of analogical word formation (הֵיקֶשׁ) available in the language of their execution; others arose from the contact between written Hebrew and the spoken vernacular, or from the influence of a source language upon its Hebrew translation. A description of written Hebrew should include the different periods, places, and styles in which it was written. Each had its own attitude to the original sources; some were sparing in innovation, others rich in additions; some preserved words as they found them, others changed both form and meaning (whether intentionally or not). The different languages with which Hebrew came in contact must also be discussed: the Aramaic of the Midrashim and the Talmud which at the beginning of the period wielded its influence as a spoken vernacular and at the end of the period as a written language which stayed alive as the vehicle for study of the Babylonian Talmud (and to some extent of the Zohar); Arabic, from the period of the ge'onim; Middle High German which had considerable influence on the language of the Jews of Germany; there are even signs of French influence (e.g., in the language of Rashi) and Italian (a little in Megillat Aḥima'aẓ, 1504, and rather more in the language of *Immanuel of Rome). There is a strong connection between the form which written Hebrew took and the nature of the culture and society which supported it. The language of poetry in Spain flourished against the background of the Golden Age of Spanish culture, in imitation of the craft of Arabic poetry. The Midrashic folk language in Germany, unaffected by the rigors of syntax and grammatical rules, is well explained by the humble character of this Jewish community which was influenced by the liturgy and the halakhah of Ereẓ Israel, took hardly any interest in science and grammar, and lacked any social or cultural environment advanced enough to provide a model for literary creation.

The Hebrew language will be described mainly, but not solely, by reference to the language of prominent figures in the world of literature or Jewish intellectual life. An account will also be given of the link between the ideas of the grammarians and the writing of good Hebrew. Nothing will be said of the pronunciations of Hebrew (for phonological developments, see *Pronunciations of Hebrew).

THE LANGUAGE OF PIYYUṬ

The first revival of Hebrew after its extinction as a spoken vernacular was in the piyyuṭ in Ereẓ Israel, where there was a considerable return to written Hebrew, not only as a language from which to quote but as a linguistic activity aimed at increasing the vocabulary with newly derived nouns and verbs. The piyyuṭim were religious poems used as prayers in public worship. Some scholars have placed the beginning of liturgical poetry as early as the third or fourth century (J. Schirmann); others have put forward later dates. The generally accepted opinion is that they date from the fifth–sixth century, in Ereẓ Israel, and were written against a background of Midrashim and spoken Aramaic. The piyyuṭim are a blend of Biblical Hebrew, eminently suitable for ceremonial religious poetry of a national character, and Mishnaic Hebrew, without which it would have been impossible to give them the homiletic, midrashic content which is their main characteristic. Zunz seems to have been the first to name the piyyuṭim "Midrashim in the guise of poetry," and it is customary nowadays to emphasize that they are versified homilies (e.g., Mirsky). The linguistic blend is apparent not only in the choice of vocabulary but also in the grammar. The extensive revival of verbs in binyan puʿal, the co-occurrence of short and long tense forms, the use of the absolute infinitive and to a limited extent of the conversive waw are typical of Biblical Hebrew; the use of binyan nitpaʿal, and complex infinitive forms like מלקטל, etc., derive from Mishnaic Hebrew.

The unique feature of the language of the piyyuṭim, however is not the blend of Biblical and Mishnaic Hebrew but its particular variety of linguistic innovation. Though the verse of the earliest known payṭanim, *Yose b. Yose and *Yannai, was not overcharged with difficult words in unusual declensions, the language of all the payṭanim was customarily referred to as ʾaẓ qoẓeẓ after the piyyuṭ read on Purim (a qerobah to Parshat Zakhor by Eleazar *Kallir):

Hebrew Language

קְצוּצַי לְקַצֵּץ אָץ קוֹצֵץ בֶּן קוֹצֵץ
רְצוּצַי לְרַצֵּץ בְּדִבּוּר מְפוֹצֵץ
פֻּלַּץ וְנִתְלוֹצֵץ לֵץ בְּבוֹא לְלוֹצֵץ
כְּנֵץ עַל צִפּוֹר לְנַצֵּץ כְּעָץ מְחַצְּצִים לְחַצֵּץ
The evil man, son of an evil man, ran
to cut down my persecuted ones;
With slander my broken ones to destroy;
The evil one when he came to do evil
was destroyed and the evil was done to him,
When he advised to shoot the shooters
like a hawk upon a bird to prey.

i.e., Haman, son of Hamdata, ran to cut down the Jews, to destroy them with slander. When the evil man came to do his evil deed he himself was destroyed and the evil was done to him; when he gave counsel that Israel be shot with arrows.

Many typical features of the piyyuṭim are indeed to be found in this poem: allusive phrases (קוֹצֵץ בֶּן קוֹצֵץ, also in the Midrash, מְחַצְּצִים), innovations in verb forms (פֻּלַּץ) and forms like כְּעָץ which has both עָץ for יָעץ and כְּ before an inflected verb. The poem demands explication, not only linguistically but as a riddle, with its reminiscences of the Bible and the Midrashim, its brevity, and its wealth of allusive phrases. *Saadiah Gaon, whose language has much in common with the writings of the liturgical poets, was aware that the language of some of the piyyuṭim was faulty (see his introduction to the Agron (more correctly Egron) and his note to his siddur, p. 225). The main critic was Abraham *Ibn Ezra who described the language of Kallir as "a breached city, with no walls" (commentary to Eccles. 5:1) and said of the liturgical poets in general that "they do not know how to speak correctly, they strive to use hard words, and say תַּחַן instead of תְּחִנָּה" (in his book Safah Berurah). There were also many other harsh critics.

The following were the characteristic features of liturgical poetry:

(a) The method of creating verb forms; what has been called "one rule for all conjugations, defective and reduplicative." This was the most highly criticized feature (Zulai, The Liturgical School of Poetry of Saadia Gaon, p. 7, bibl.). Examples are עָשׂ for עָשָׂה (conjugating a ל״ה verb as though it were סָע ,(ע״ו for ע״ו) נָסַע for עָץ ,(פ״נ for ע״ו) יָעַץ for פ״י) and so on. However, this mixture is only found in the perfect. Forms like יָעוּשׂ (the imperfect of עושׂ if it were an ע״ו verb) and יָסוּעַ do not occur. Even in the perfect most of the forms can be explained simply as deletion of the first (סַע from נָסָע) or last (עָשׂ from עָשָׂה) letter; only a few isolated forms show a real conversion to a different conjugation, usually to ע״ו, in other forms of the perfect, e.g. חַזְתָּה ,פַּצְתָּה. A more plausible approach, therefore, is to describe this method of conjugating verbs (which was not explained by the liturgical poets themselves) in the terms used by Saadiah Gaon in the second chapter of his Sefer Ẓaḥot, as deletion of the initial or final letter of a particular form, and not necessarily of the root consonants, by analogy with certain biblical forms: סָע after עָשׂ ,יִסַּע after וַיַּעַשׂ and so on. The deletion of the nun is thus not conditioned by šewa naḥ, nor is the deletion of the he by the shortened imperfect form. Forms such as תֶּחִי ,מַעַשׂ, and תַּאַו can similarly be explained as deletion, by analogy with מַעַל־מַעְלָה in the Bible. With the recognition of the tri-consonantal basis of the Hebrew root (by *Ḥayyuj) such forms were strongly criticized by grammarians and poets in Spain, and this technique, much used by Saadiah Gaon, almost disappeared from secular poetry after the period of the payṭanim. (N.B. such forms as פָּץ ,בָּט, etc. are not evidence that the weak roots were considered biconsonantal. Menahem and Dunash, who were unaware of the tri-consonantal nature of such conjugations, established the root שב for verb forms like יָשַׁב ,שׁוּב and נָשַׁב, yet did not mix the conjugations in their writings, and derived the forms in accordance with biblical use. They never substituted קמית for קַמְתָּ for example).

(b) The liturgical poets created many new words by conjugating roots in all the binyanim; for them the binyan was part of the automatic inflectional system, like tense and person. The later poets of Spain saw the binyan as non-automatic, not subject to unrestricted analogical extension, confined to forms found in biblical Hebrew and mishnaic Hebrew. Not so the writers of the piyyuṭim. They derived many words from nouns: סִמְדֵּר (from זִלְעֵף ,(סְמָדַר (from הִבְטִין ,(זַלְעָפָה (from בֶּטֶן), and even from adverbs and particles: טֵרְמַנִי from טֶרֶם and לְבַלְעֵד from בִּלְעֲדֵי. They also incorporated derivational affixes and suffixes as root letters: לְהַתְשִׂיר from תְּשׁוּרָה by analogy with mishnaic לְהַתְרִים from תְּרוּמָה. This abundance of morphological variation is the hallmark of liturgical poetry, and has been vilified as bizarre by its critics. The grammarians and some of the poets in Spain rejected the alternate forms because they "changed the holy writ" (Dunash ben Labrat's reply to Saadiah Gaon, 95), stating "we shall read every word in the form in which we found it" (ibid.). Abraham Ibn Ezra declares: "a man must use a word in the form in which it is found" (Moznayim, 33). (A similar statement was made by Moses Ibn Ezra in Širat Yisrael, 148). Both Abraham and Moses Ibn Ezra criticized the use of given verbs in binyanim in which they did not occur in the Bible. Dunash denounced a change in noun forms but allowed the use of different binyanim. The early critics rejected these changes because they wanted to preserve biblical Hebrew; later critics deprecated poets who adopted different forms and made innovations which did not contribute to the sense of the poem.

Criticism of the language of piyyuṭ is intrinsically criticism of its style. Abraham Ibn Ezra mentions four flaws – two concerning the language and the other two content and poetic devices: the piyyuṭ is (1) influenced by a foreign tongue, i.e., the Aramaic of the Talmud; full of (2) grammatical errors; (3) riddles and fables obscure in meaning; and (4) homilies. In comparison to the prayers, the piyyuṭ is obscure in language and style and is unfit to be used in liturgy (Comm. Eccles. 5:1). Basically, however, the language of the piyyut is difficult because of its many allusions: עֲמוּסִים ("the encumbered") standing for "the Children of Israel," אֵיתָן ("the strong") for "Abraham," etc. Many of the forms, drawn from the Midrashim, were more difficult for the Jews of Spain than those of Ereẓ Israel who knew the Midrashim well. Graetz sharply criticized the piyyuṭ, while Samuel David Luzzatto explicitly defended Ha-Kallir: "not because of ignorance and duress did he write it so, but to embellish the style," (introd. to Maḥzor Roma, 1861), "Eleazar Kallir was not tongue-tied but for his wisdom and of his own will did he write it" (Letter to S.J. Rapaport, 1884). Zunz described the language of the piyyuṭ with great understanding and he knew that it was written for the taste of that generation (Ha-Derašot be-Yisra'el, pp. 184–5). In his opinion the plenitude of vocabulary serves as an ornament of style, and many forms are simply "nonce words," not meant to be established as part of the language (see bibl.).

The liturgical poets left nothing in writing which would inform us of their view of language and style, but it is clear that they regarded an active, prolific use of derivational inflections as one of the glories of the language. The wealth of forms is similar to the technique of listing synonyms (the liturgical poets liked to fill a line with a long list of synonyms or near-synonyms), the use of word play, and the use of a recurrent rhyme word. The invention of new forms which do not carry new meanings creates both a richness of sound and a degree of synonymy which are among the rhetorical techniques of liturgical poetry. For Bialik also, criticism of the literary value of the piyyuṭim is inseparable from criticism of their language: "the period of Ha Kallir and his disciples was a time of infatuation with liturgical poetry, which became more and more sentimental" (Širatenu ha-Ẓeʿira). He condemned the "makers of acrostics" and the "tasteless stammerers" whose language was "like the gravel (אבני חצץ) of aẓ qoẓeẓ."

(c) The letter kap (-כ standing for כַּאֲשֶׁר as well as -כ for כְּמוֹ) can be prefixed to an inflected verb: לוֹ לָיְלָהגֵּר צֶדֶק נִצַּחְתּוֹ כָּנֶחֶלַק "you made the righteous proselyte prevail at midnight" (Yannai in "Wa-Yehi ba-Ḥaẓi ha-Laylah"). So also כְּיֵלֵךְ ,כְּדִבַּרְתָּ ,כְּהָלְכוּ. This is a characteristic feature of liturgical poetry in Ereẓ Israel. Dunash regarded the prefixing of כ to verbs in the perfect tense as a rule of analogy from biblical practice (הֶהָלְכוּא, Josh. 10:24 – replies to Saadiah Gaon, 114). In his poem against *Menahem ibn Saruq he uses the form כְּשָׁקַט. Although Ibn Janaḥ permitted the use of kap before a perfect tense for reasons of scansion (to provide an iamb (Harikma 45–46)) it was not used in secular poetry in Spain, undoubtedly due to the influence of the grammarians. Saadiah Gaon used kap before perfect tenses, but in Sefer Ẓaḥot, written late in life, he denounced its usage. Examples of kap before perfect tenses are also to be found in the writings of Hai Gaon. However, in those communities which drew their inspiration from the Midrashim, the halakhah and the liturgical poetry of Ereẓ Israel, this linguistic feature continues to occur, and not only in acrostics. In Megillat Aḥima'aẓ (Italy, 1054) we find כְּבָאוּ ,כְּשָׁמְעוּ, and in the German elegies and poems written about the horrors of the Crusades we find כְּהָלַךְ and כְּפָחֲזוּ.

From the Spanish period until the Enlightenment, the liturgical poets were charged with ignorance of grammar. It would be going too far to say that they had no understanding of language; their view of word creation is not in line with accepted grammar. A realization of the motives which led to such abundance of morphological-stylistic innovation can bring us closer to understanding the liturgical poet as a deliberate, if inartistic, manipulator of language. In recent years new light has been thrown on the language of the piyyuṭim as a linguistic and not merely a stylistic phenomenon. Study of Palestinian Aramaic and the language of Hebrew Midrashim written in Ereẓ Israel has revealed that the language of the piyyuṭim is based on "Palestinian idiom" (Zulai). Words thought to be arbitrary innovations invented by the liturgical poets have been shown to be rooted in the language of the Midrashim and the Targum. Yallon has pointed out words which passed from the Midrashim to the piyyuṭim (גהר meaning reproof, פנה meaning look, יאש meaning weak, לִבֵּב meaning shout, ברור meaning strong and existing, and others). Expressions from the Hebrew spoken in Ereẓ Israel survived in the piyyuṭim: קְפִידַת רוּחַ – severity – and זְרִיקַת תְּפִלּוֹת – prayer (noted by Zulai), or מוֹדַע for תְּעוּדָה document, and מַסְפֵּק – danger (Shalom Spiegel). S. Lieberman has also pointed out the affinity to midrashic language (see bibl.). The language of Ereẓ Israel can elucidate difficult passages in liturgical poetry, and makes it clear how the creation of the poets was natural and not artificial. A comprehensive description of the language of the piyyuṭim against the background of the languages of Palestine and the attitudes to grammar out of which it took shape would contribute greatly to a clearer understanding of this first important manifestation of written Hebrew.

SAADIAH GAON

It was Saadiah Gaon who brought about the great revival of Hebrew writing in Babylon. Actually, even before his time the use of Hebrew in writing had not been set aside completely. In Ereẓ Israel there had been liturgical poetry, Midrashim and collections of legal decisions, the best known, Sefer ha-Maʿasim ("The Book of Court Cases" or "Judgments"), was collected at the beginning of the geonic period.

After the period of Saadiah Gaon the Palestinian geonim continued to write a good Hebrew, and Ibn Janaḥ affirms that "the men of Tiberias excelled all others in the purity of their Hebrew." In the talmudic academy of Damascus, which took over from that of Ereẓ Israel, halakhah and metrical, rhymed prose were written in Hebrew (in the 11th century). The same kind of thing occurred in Babylon. In the Talmud short extracts from the amoraim, consisting of a presentation of the problem and a brief discussion, are written in Hebrew; detailed discussion in Aramaic comes later. At the end of the eighth century Pirkoi ben Baboi wrote chapters of halakhah in good Hebrew. The collections of halakhic decisions, Halakhot Pesukot of Yehudai Gaon and Halakhot Gedolot of Simeon Kayyara both contain sections in Hebrew. Needless to say they derive from Palestinian literature, but they do bear witness to the fact that even in Babylon, Hebrew had not given way completely to Aramaic, and was used for special purposes, e.g., for halakhic decisions. Several Aramaic books on halakhah were translated, or translated and edited in Hebrew, notably Halakhot Re'u to Halakhot Pesukot and the book We-Hizhir to the responsa of Aḥai of Shabḥa. Conventional opinion (Poznansky, Epstein, Assaf, and Ginzberg) holds that the translations were done in Ereẓ Israel, or at the very least in Greece and Italy, since Hebrew translation could only have been carried out where Babylonian Aramaic was unknown. Nevertheless, according to S. Abramson it is quite possible that they were written in Babylon when Aramaic had given way to Arabic. Therefore the Halakhot Pesukot, for instance, was translated into Arabic too. Linguistic features regarded as typically Palestinian – אָדָן for לוֹכַל ,אָדָם as the infinitive of אכל – cannot therefore be taken as evidence of the place of composition; the translators in Babylon could well have considered language of Ereẓ Israel a fitting model for good Hebrew. If they were written in Babylon, it could not have been earlier than the time of Saadiah Gaon.

Saadiah Gaon introduced the writing of liturgical poetry in Hebrew into Babylon. He was followed by *Hai Gaon, whose language is generally simple, but very similar in its techniques of word creation, patterns, and usage to Saadiah's language. Saadiah Gaon brought a consciousness of the need for beauty to the writing of Hebrew. In the introduction to his dictionary, the Agron, he writes of Hebrew as a woman who had been slighted when the Children of Israel preferred the imperfect foreign tongues of exile to her own beauty of expression. The Agron was designed to fashion Hebrew into a proper instrument for the writing of poetry. It is commonly held that the language of Saadiah Gaon is a link in the chain connecting the language of the piyyuṭim and the language of Spanish Jewish poetry; this view is expressed primarily in "The Liturgical School of Poets of Saadiah Gaon" by Menahem Zulai. The language of Saadiah Gaon is far removed from the biblical purism of Spanish Jewish poetry, as we shall see. However, though it is true that it shares features with the language of the piyyuṭim, and continues this tradition, it also foreshadows in several ways the approach to language of the Spanish poets. Saadiah Gaon wrote piyyuṭim in Hebrew as well as polemical literature Essa Meshali, Sefer ha-Galuy and halakhah (Sefer ha-Moʿadim). He wrote an introduction to his Agron, his grammar of Hebrew (Sefer ẓaḥot), philosophy (Emunot ve-Deʿot) and responsa in Arabic. It was he who initiated this duality in Jewish writing in the Middle Ages: Hebrew for poetry, and Arabic for prose, even for those who honored Hebrew, the mistress, more than Arabic, the serving maid (expression of Al-Ḥarizi in his book Taḥkemoni and Solomon ibn Gabirol in Ha-ʿAnaq).

It was Saadiah Gaon who introduced the concept of "pure" language – zaḥot ha-lašon – to Hebrew writings and grammar (on the basis of Isa. 32:4), thereby creating a Hebrew cognate of the Arabic faṣāḥa which is also etymologically related to the Arabic term taṣḥīḥ. "Pure" language for Saadiah Gaon is a linguistic ideal, beautiful, clear, and correct, with all forms derived according to proper rules of analogical formation, free from errors of irregular word formation. Analogy (הֵיקֶשׁ) is permitted to operate according to biblical patterns which, in his opinion, were "fertile" but not according to infertile patterns. This view of purity of language matches the primary concept of faṣāḥa. Later, "purity" will be able to express various linguistic and stylistic qualitative features that affect poetic ornamentation and rhetorical figures (such as plays upon words and synonyms). According to Saadiah "purity" of language is not just the passive use of received vocabulary; it welcomes innovation, since it is linguistic activity that shows knowledge of the language and makes it beautiful. Thus Saadiah Gaon used Hebrew like the liturgical poets who preceded him. He created new verbs by using existing roots in all the binyanim, each with its special meaning, e.g., מַקְוֶה – giving hope, הַנְפִּישׁ gave rest (לְהַחְדִּיל ,(נֹפֶשׁ (from חָדַל) and so on. He formed verbs from nouns: הִתְהִים (from המְעִין ,(תְּהוֹם (from מָעוֹן).

In deverbal derivation Saadiah used the various nominal patterns which he regarded as nouns of action (infinitives), though not other noun-forms: לַעַט ,טְפִיפָה ,דִּרְשׁוֹן ,שִׂטְמָה (hatred) these are the commonest types – and also מִדְבָּק (for דְּבִיקָה) מִגְלַעַת), "quarrel"), מַחֲמׁרֶת ("restriction" for חוּמְרָה). Like the liturgical poets he changed the form of extant words without changing the meaning, as he himself stated explicitly (with respect to word expansion and deletion in the second section of Sefer Ẓaḥot, and with respect to the variations in the form of nouns derived from verbs, see replies of Dunash, 122): wa al-maʿnā wāḥid "the meaning is one," i.e., the different forms are "equivalent in meaning," differing only in "articulation."

(a) He used expanded forms of words in accordance with techniques which he explained in his grammatical writings (he referred to expansion as tafkim): סִדְסֵד for דּוֹתֵת ,יָסַד (from לְדוֹגֵג ,(דָּת for תִּלְתֵּל ,לָדוּג for סתסת ,תָּלָה for אימימה ,סתת for אימים ("horror"). Lengthened forms of the imperative and the imperfect like those of the infinitives are used without any implications of modality: יָבִיאָה ,יוֹצִיאָה ,יַעֲלֹזָה ,לְנַשְׂאָה.

(b) He omitted letters in various word forms, on the model of contractions found in the Bible: עָט ,בָּט ,פָּץ ,סָע – shortened perfects; לְהַעַל – shortened infinitive; מַעַל for מַעֲלֶה – shortened present participle; תְּחִיָּה – תֶּחִי ,חֵזֶה – חֵז ,מַעֲשֶׂה – מַעַשׂ – shortened verbal nouns; and of course many shortened imperfects – תַּהג ,(תִּ)תְאו – since this form does not carry any jussive meanings and can be used to form further shortened forms without any effect at all on the shade of meaning.

(c) He used alternative forms of words by analogy with doublets found in the Bible: עָרִיס for עֲרִיסָה ("dough" like כפור – כפורת ,יסוד – יסודת ;(גְּלִילָה – גָּלִיל (like חֶפְצוֹן ;(יבוש – יְבֹשֶׁת (from חֵפֶץ, like חרבון from חֹרב) and so on. The explanations given earlier for the word creation of the liturgical poets are made explicit in the writings of Saadiah Gaon. All these derived forms, he says, come easily to the language; they are mere changes in form which do not necessitate any special shade of meaning. In Saadiah Gaon's opinion, the principal source for Hebrew writing is the Bible, though he also made no small use of mishnaic Hebrew. Choice was dictated by the needs of style. He did not think of biblical Hebrew and mishnaic Hebrew as separate entities; the latter simply completed the documentation of the words in the former. Like the Arab grammarians of his time, he lacked any historical sense of earlier and later periods in the development of the language. He thought that all the words in the Bible were fit for use, including the rarest and oddest; further words could only be created by the operation of analogy on this vocabulary. He continued the tradition of the liturgical poets in his use of allusive phrases: בְּנֵי אֶלְעָד ("children of eternity") for the Children of Israel and פְּרִיזָה for the Temple and many others. Liturgical poetry was a source of literary inspiration for Saadiah Gaon in many ways: specific usages, nonce-words, allusive phrases, rhyme and alphabetical arrangement, and such words and expressions as גַּיְא ,נֶשֶׁם ,פֶּצַח ,צֶרַח ("valley," i.e., land), יָחִיד ("unique," i.e., Isaac), גַּבְנֹן ("peak," i.e., Mt. Sinai). Liturgical poetry was a source of style and thematic material, but not of linguistic innovation per se; he would not accept any new forms which did not satisfy his own linguistic principles. Such words, he thought, could be invented by anyone who knew the language and had the inclination. The main features he shares with the liturgical poets is his constant use of derivation without change of meaning. In the following respects he foreshadows the language of Jewish poetry in Spain

(1) There was a close relation between his linguistic inventiveness and his views on language; like the poets of Spain he remained faithful to the rules of grammar (though since his conception of the language was different, the results are also different).

(2) It was he who initiated the criticism of the language of the piyyuṭim (in the Arabic introduction of the Agron, and Siddur p. 225).

(3) Though he continued to use the techniques of the liturgical poets, he did so with reservations. He derived nouns according to the patterns of verbal nouns only, and placed restrictions of the freedom to create new words by analogy. (In Spain also, vide infra, innovations were more frequent with verbs than with nouns). Therefore like the Spanish poets he made considerable use of participles, active and passive, as adjectives (פּוֹתֵל – "crooked," כָּחוּד – "hidden").

(4) He used short and long tense forms freely, without any distinctive modal significance, and this lack of specific meaning facilitated morphological innovation by analogy. The Spanish poets did the same (vide infra).

(5) The rules of analogy are binding as far as the derivational inflections are concerned, but a word may have many different shades of meanings. Saadiah Gaon did not always use a word in the sense in which he translated it in his translation of the Bible. He translated תֶּבֶל as "punishment" (in Ar. Dāhiya) and then uses תַּבְלִית – by analogy with חֶרֶס – חַרְסִית (Siddur Sa'adyah Ga'on p. 198) – in the sense of "abomination." He translates אֲרֶשֶׁת שְׂפָתַיִם in Psalms 21:3, as "permission" (Ar. isti'dhāna) and explains this interpretation by reference to Ezra 3:7; in his poetry he uses אוֹרֵשׁ and רושֶׁה as synonyms, with the meaning "say." The technique of allusive phrases also depends on the view of language that a word may have many meanings, all available for use.

(6) Saadiah Gaon had his own opinions as to which were the proper patterns for analogical word formation, and the frequency of a word had no relevance. A typically mishnaic word which occurs only once in the Bible is considered biblical. The Spanish poets took the same stand.

A conventional explanation of the language of liturgical poetry holds that it was not difficult for its audience, who were learned in the Midrashim. Saadiah Gaon certainly thought the piyyuṭim difficult to understand, and regretted that most people had a scanty knowledge of the language. He thought that people liked piyyuṭim even though they did not understand them (Siddur Sa'adyah Ga'on p. 156). Prayers and entreaties (בַּקָּשׁוֹת), by which a man might draw near to his Creator, were written by Saadiah Gaon in language devoid of allusions and morphological innovations, lest the language mar the prayers and therefore were praised by Abraham Ibn Ezra (commentary to Eccles. 5.1.). True innovations are scarce in the language of Saadiah Gaon, and occur not in his poetry but when he needs to coin technical terms; for these, like the Spanish writers (infra) he uses the method of loan-translation. He borrows both the concept and the way of expressing it: אגרון (lit., "hoard") for dictionary, a loan-translation of the Arabic ǧamhara, (a verbal noun). There are other synonyms for dictionary in Arabic from the roots ǧml, ǧmʿ and ʾiḥatawā, all meaning "hoard" or "collect" יסוד, from Ar. ʾaṣl (root).

Arabic Influence

The influence of Arabic is more strongly felt in his views on language than in his actual grammatical innovations. His grammatical theory is strongly influenced by the opinions of Arab grammarians on analogy, e.g., in his abundant use of the קֶטֶל pattern (as most fertile fiʿl, faʿl, and fuʿl in Arabic) in not differentiating between what would now be called the infinitive (קָטוֹל) and the verbal noun (קְטִילָה). There are very few Arabicisms in his language, even fewer than in the poetry of Spain (infra): צרחת בה – the root צרח is common in liturgical poetry, but the meaning "declare" and the use of the preposition ב are from Arabic עֵץ נִגְדַעַת (in Tešubot ʿal-Hiwwi 4) where for reasons of rhyme the word עֵץ ("tree") is construed as feminine, as in Arabic. In his translation into Arabic he preferred words which were alike in sound to the Hebrew, and sometimes did the same in his own writing. In the introduction to the Agron he uses the expression חוֹדֶרֶת for "woman," from the expression יושֶׁבֶת פְּנִימָה בַּחֲדָרִים. This is close to נָוָה which means "wife" in the writings of Saadiah Gaon and is etymologically similar to the Arabic al-Muḵdara (girl kept indoors).

SPANISH HEBREW POETRY

Spanish Jewry followed the spiritual center that was in Babylonia as far as the halakhah was concerned; however in literary writing an important innovation took place there in comparison to both Ereẓ Israel and Babylonia. The duality – Hebrew for poetry and Arabic for prose – which started with Saadiah Gaon was fulfilled to a large extent in the literary activity of the Jews of Spain. Arabic replaced Aramaic as the vehicle for non-poetic expression (mainly halakhah) and became the language for prose writing (grammar, medicine, philosophy, exegesis, etc.) although there were still scientific books written in Hebrew (see below). Hebrew was used for poetry although some secular poetry was written in Arabic and Aramaic.

The Hebrew of poetry in Spain underwent a fundamental change when secular poetry became a separate and respectable literary genre. The beginnings of secular poetry are to be found in the polemic writings of Saadiah Gaon (which are the forerunners of poems of personal quarrels and denigration) and secular poetry became an accepted art – important and widespread in Spain – starting with the wine and war poems of Samuel ha-Nagid (d. 1055).

The earliest liturgical poets in Spain (especially Isaac ibn Ghayyat, Ibn Abitur, and Ibn Khalfon) drew their linguistic and stylistic inspiration from the piyyutim; they wrote mainly sacred poetry, in the same style and language as other liturgical poets, uninfluenced by the Arabs as the Arabs had no religious poetry. But secular poetry, a personal art (unlike religious poetry which was designed for public worship), was the product of Arab culture, and took shape in the image of Arabic poetry with which it competed by imitation. Liturgical poetry could not provide suitable vehicles for the writing of secular poetry. Linguistic change actually crystallized in secular poetry, under the influence of Arabic; iambic meter was taken over, and there is a close tie between the meter and the formation of words and verbal conjugations. In religious poetry it was used only sparingly; Keter Malḵut by Solomon ibn Gabriol, for example, is written in one of the meters of liturgical poetry, with a fixed number of syllables per line. The move towards biblical purism, as understood by the poets of the Middle Ages, began in secular poetry.

The choice of Hebrew for poetry and Arabic for prose is closely interlinked with the move towards a biblical Hebrew. The writing of secular poetry in Hebrew was supported by the continued writing of religious poetry in Hebrew (which there was no cause to write in Arabic). The linguistic duality had its counterparts in the surrounding culture: old Spanish for speech and Latin for writing in Christian Spain, Andalusian Arabic for speech and the Classical Arabic of the Koran for poetry in Muslim Spain. The Jewish poet would rather make a careful, diligent, accurate study of the language of the Bible than learn how to write the language of the Koran. He had to choose between two languages which both required considerable study. (Prose was written in Middle Arabic, which did not need special study.) Perhaps, however, the fact that it suited his background is of little account compared with the national and religious feeling that biblical Hebrew had a special status, was a "superior language," a "very choice tongue" (Solomon ibn Gabirol in Sefer ha ʿAnaq) and "a wondrous language" the best of all the languages, and of course richer and more beautiful than Arabic (Al-Ḥarizi in the first chapter of Taḥkemoni). See also the speech of the companion in Judah Halevi's Sefer ha-Kuzari, Part 2, 68.

Biblical Hebrew was extolled as a "pure" language by no means inferior and indeed superior to the "pure Arabic" which was used for poetry. It was eminently suitable for the writing of verse comparable with Arabic verse, since it had similar virtues: (a) The similes, metaphors, and other figures of speech in the Bible were well suited to poetic style. In the last chapters of his book Širat Yisra'el, Moses Ibn Ezra quotes examples from the Bible for every one of the rhetorical figures used in Arabic poetry. (b) The tradition of a fixed vocalization could serve as a basis for the iambic measures introduced from Arabic poetry. (c) The study of the grammar of biblical Hebrew was highly developed in the Middle Ages; there were few references to features of mishnaic Hebrew and the Hebrew of the period in grammar books. Arabic poetic language was also subjected to perpetual scrutiny by Arab scholars. (d) Biblical commentary added to the vocabulary a wealth of meanings and shades of meaning which were essential to a richness for poetic expression.

The introduction of a wide range of subjects – passion and wine, war, dispute, lampoon and jest, elegy, panegyric and self-aggrandizement, love, friendship and marriage – was accompanied by the use of iambic meters. Instead of every line containing a fixed number of syllables, there was a regular alternation of full vowels and šewa naʿ, i.e., reduced vowels. Dunash ben Labrat is usually considered the first to introduce iambic meters into Hebrew poetry. He used "well-scanned newly invented distinguished metrically constrained poetic forms" (Replies of Dunash to Menahem 4, 19). And this view is confirmed by the accusations leveled against Dunash by the disciples of Menahem, that he abused the forms of the language: "our holy tongue destroyed and left it null and void, because he had employed a foreign measure" (Replies of the disciples of Menahem 7, 44.) Menahem's disciples themselves phrased their replies in iambic measures, to show that their condemnation did not stem from poetic incompetence. The iamb was soon accepted by all poets as the proper measure for secular poetry. In the beginning it was difficult to adjust to the new meter; Al-Ḥarizi said of the language of the period when the iamb was first introduced that "the writers of the time wrote bad measures." It was the meter which usually determined the choice of words: שֶ or אָז ,אֲשֶׁר or, אֲנִי ,אֲזַי or אָנֹכִי, depending on the needs of the rhythm. Sometimes it even led to a change in the basic form of the word (e.g., קְרַב, with a šewa, for קֶרֶב); Ibn Janaḥ realized that the exigencies of meter could open the way to deviations from proper inflection (Ha-Riqma, 226–7). Many long and short tense forms were chosen to fit the meter (imperfect forms ending in וּן, e.g., יְרִיבוּן for יָרִיבוּ, were very useful) and rare words became common because they could provide iambs. לְמַעַן (from Neh. 6:13) or בְּיַעַן (which occurs only once in the Bible, in the combination יַעַן בְּיַעַן) were regularly used in place of יַעַן. The letter he with a šewa at the beginning of a word for emphasis (הֲכִי, "indeed"; הֲלִי, "indeed to me") fitted the meter and was grammatically acceptable (Ibn Janaḥ, Ha-Riqma, 68: "he to establish or verify a fact").

"Pure" language is, above all, grammatically accurate, and Spanish Hebrew poetry, especially secular poetry, is characterized by the poet's strict adherence to the rules of grammar (an approach which, as has been pointed out above, begins with Saadiah Gaon). In his Sefer ha-Riqma Ibn Janaḥ notes linguistic usages of the poets and affirms that a poet should not be blamed for linguistic deviations necessitated by the requirements of poetic forms (pp. 226–7 and p. 275); it even happened that poems were corrected by their readers according to the rules of grammar (Ha-Riqma, p. 275). In his book Širat Yisra'el, Moses Ibn Ezra teaches the art of writing poetry. Matters of grammar, which "add salt to the food," are explained first, before any discussion of decorative figures, and the poet is told which grammar books are worthy of study before he is referred to any books on prosody (p. 100). Poets who "did not follow the grammarians" are condemned (p. 65) see also Al-Harizi in Tahkemoni ch. 18). The ideal form of poetic language is given full expression in Širat Yisra'el from which the above are quotations. Essentially it is a matter of adherence to all the rules of the grammar of biblical Hebrew, with no innovations in form due to analogy, since "the language must be imitated, but without creating new words" (147). Verbs must not be used except in the binyanim in which they occur in the Bible (148). The given form of a word must not be changed. Care must be taken not to turn masculine into feminine or singular into plural, or vice versa. However, according to his system it is permissible to create new forms in the infinitive, קטול (149), and it follows from the general trend of his remarks (though he does not say so explicitly) that a verb could be inflected in all the forms of the given binyan: short and long tense forms, and hip̱ʿil on the model of יְהַכִין, which was convenient for iambic meter. The ban on analogical word formation is a reaction against the copious use of such forms by the liturgical poets. (A similar opinion to Moses Ibn Ezra's can be seen in Abraham Ibn Ezra's Ẓaḥut, 26, with reference to the ban on analogy for nouns, by contrast with verbs and in the replies of Dunash to Saadiah Gaon, no. 95). However, there was not a single poet who abstained completely from morphological innovation, and even Moses Ibn Ezra allowed himself a few new forms, as he himself admitted (p. 156–7). In his opinion, his language was tainted with error because of human weakness, and lack of skill in his early poetry.

Moses Ibn Ezra's deviation into analogy are few: תְּבוּנָיו (the Bible has the form עֲלוּמוֹתַי ,(תְּבוּנוֹת (for מִגְדָּנַי ,(עֲלוּמַי (for מַעֲצָב ,(מִגְדְּנוֹתַי (for מַעֲצֵבָה or עֶצֶּב in the Bible), and גַּחַל ,אַהַב singular forms derived from גַּחֲלֵי ,אֲהָבִים – though incorrectly since the correct singular forms occurring in the Bible are אַהֲבָה and גַּחֶלֶת. Moses Ibn Ezra's purity – ẓaḥot – of biblical language and abstention from morphological innovation are maintained to a fairly similar extent by Judah Halevi and Abraham Ibn Ezra. However in their work there are also exceptions, giving every secular poem a flavor of liturgical language. A further feature the secular lyrics share with the piyyuṭim is the use of allusive phrases: יְחִידָה (unique) for soul, צִיר (messenger) for Moses, יְקוּתִיאֵל for Moses, and so on. Considerable use of word derivation is made by Samuel ha-Nagid, who wrote when the language of Spanish poetry was just beginning to take shape (שִׂבְרָה חֲרִיזָה ,אקדָחה (גֶוַע אקדָח ("hope"), and innovations in the binyanim. Indeed in his own time he was censured for his use of analogical word formation, especially in the creation of nouns (Širat Yisrael, 67). Solomon ibn Gabirol also has many new forms not found in the Bible (David Yellin counted 1,500), and though most are found in his sacred works, there are some in his secular poetry: בְּהִיקָה ,תַּחַןחֶפְשׂוֹן, etc. The most serious offense in the eyes of Moses Ibn Ezra was the creation of verbs and adjectives from nouns; no wonder he criticized Solomon ibn Gabirol for such forms as מְשֹׁהָמָה ,מְיֻשָּׁפָה and פְּנִינִיָּה (Širat Yisra'el 151).

Most violations of the ban on analogy are to be found in the binyanim; it is more difficult to avoid using verbs in active and passive conjugations than to refrain from inventing new noun forms. The use of passive conjugations – puʿal and hopʿal – is particularly common in the language of poetry, partly under the influence of Arabic. Various forms of verbal nouns were fashioned – חֶפְצוֹן ,גֶּוַע ,בְּהִיקָה, – and many participles, active and passive, since this was the only way of creating adjectives (none of the adjectival patterns were productive). Singular forms were derived from plurals to a degree exceeding the limits prescribed by Moses Ibn Ezra (Al-Ḥarizi has סַנֲור from סַנְוֵרִים, and Ibn Ezra has פָּלִיל from פְּלִילִים used in the sense of "judge"). Even more frequent are plurals for singular: נְרָדִים ,לְשָׁמִים ,אַחְלָמוֹת and רְפָשִׁים (plural forms are much used in rhyming). Nevertheless, fundamentally and in comparison to liturgical poetry and to poetry of Saadiah Gaon on the one hand and to Hebrew prose literature on the other, secular poetry should be regarded as faithful to the given forms of biblical vocabulary. Innovations of language are of course far commoner in religious poetry. Certain of the techniques found in the piyyuṭim, anathema to those believers in "pure" language who hearkened to the grammarians, do occur in religious poetry: כְּ plus perfect tense (in Ibn Khalfon and quite frequently in Solomon ibn *Gabirol) and the use of "shortened perfects" such as עָט (for עָטָה) and סָט (for סָטָה) in Solomon ibn Gabirol.

Moses Ibn Ezra believed that analogical innovation marred the purity of biblical Hebrew, but sparing use of mishnaic Hebrew did no harm: "If we avail ourselves sometimes of the language of the Mishnah, this is acceptable, since its words are pure Hebrew" (p. 59). The language of secular poetry, unlike sacred poetry is free of forms typical of the piyyuṭim. Not only literary motives and content-words from the Mishnah are used, אִסּוּר ,תְּנַאי מִדְרָשׁ etc., – but also form-words such as צָרִיךְ ,פְּרָט לְ -,כָּאן. The word כְּאִלּוּ is popular, and the structure לִקְטֹל + מ (yaday kebedim mi-lesapper – Judah Halevi) is quite common. Binyan nitpaʿal appears sporadically (Abraham Ibn Ezra niṭrapeʾta), and not because of acrostic composition; Moses Ibn Ezra writes יִשְׁתַּבַּח in place of the biblical יְהֻלַּל. Samuel Ha-Nagid, a great talmudic scholar, especially introduced Mishnaic-Hebrew usage. At the end of the period, from the 13th century onwards, the adherence to biblical Hebrew weakened; Meshullam di Farra, for example, has more usages from the Midrash and even from the languages of his time, סְפִירָה from the kabbalists, בַּהֲמִי, etc. It should be recalled that it was not frequency and provenance of a word which determined its value. Not merely were rare words acceptable in poetry, but a word that occurred only once or twice in the Bible and regularly in the Mishnah was nevertheless regarded as biblical; it was used as commonly as a biblical word, and not sparingly like the mishnaic vocabulary. The alternation between שֶׁ and אֲשֶׁר is thus between two biblical words, and words of the קְטִילָה pattern (very common in the poetry of Samuel ha-Nagid for example), though regarded by present-day linguistic research as typically mishnaic, are treated as biblical on the strength of חֲנִינָה ,אֲכִילָה, and other biblical examples.

Despite the restraints on analogy, there were ways of diversifying the vocabulary; any binyan could be used in the long forms (לְכַפְּרָה ,יְרִיבוּן) and in the short forms (תְּשַׁו ,יִתְעַל) in exact conformity with the conventions for adding or subtracting letters found in the Bible. In consequence, and by contrast with the language of liturgical poetry, forms like עָט for עָטָה and מַחַן for מַחֲנֶה are scarcely found in Spanish poetry. The free use of lengthened and contracted imperfect forms, which are useful for rhyming and scansion, derives from the writer's belief that such changes of form, unlike changes of binyan, had no effect upon the meaning. Medieval grammarians did not interpret the lengthened imperfect as cohortative, or the shortened imperfect as jussive (see Ibn Janaḥ, Ha-Riqma p. 96). Hence the license to use such forms freely accorded with the grammatical theory of the period. And there was similar freedom to meet the stylistic demands of poetry by using the waw conversive, pausal forms, possessive suffixes on the model of מִנוּחָיְכִי ,עָלֵימוֹ, and the imperfect hipʿil forms like יְהַפְרִיד ,יְהָכִין (even Moses Ibn Ezra, though not to excess). The restrictions on analogy reduced the abundance of forms, but the language of secular poetry is rich in means of expression, since the poets gave the biblical vocabulary a wealth of meanings. The formal, grammatical features of a biblical word were binding, but not the semantic. (Once the tri-consonantal form of the root was fully established, changes in the interpretation of a word – as revealed by the dictionaries – exceed changes in the formal analysis of root and declension – as revealed in the grammar books.) Abraham Ibn Ezra and Moses Ibn Ezra, for example, insisted on the correct use of a word, in accordance with its meaning in context, and were aware that a word might have different meanings, related and quite unrelated.

The poets were well aware that a particular word had been interpreted in different ways by the lexicographers, the commentators, and the translators, and this enabled them to choose whichever meanings they required for their poetry. They used duality of meaning to rhyme a word with itself, (צִמוּד שָׁלֵם). *Al-Ḥarizi does this in Ha-ʿAnaq and Moses Ibn Ezra in his poem Ha-ʿAnaq, also known as Taršish. The latter rhymes אֵיד ("misfortune") with אֵד ("mist"), and צִיר ("pain") with צִיר ("messenger") and צִיר ("door"). Judah Halevi writes דְּלָקוּנִי אֲהָבֶיךָ וְאַחְרֶיךָ דָלַקְתִּי ("my love inflamed me and I pursued thee"), using דלק in two different senses. אַגְמוֹן (Isa. 9:30) is used by Al-Ḥarizi with the meaning "fortress" (as interpreted by *Ibn Janaḥ) and by Samuel ha-Nagid with the meaning "branch" (as translated by Saadiah Gaon). Many words may be interpreted with the aid of the medieval dictionaries, and these interpretations are supported by the biblical commentators and the translators. Moses Ibn Ezra combines שִׁוָּה ("placed") with שִׁוָּה ("straightened") and שִׁוָּה ("lied-deceived"), in accordance with Ibn Janaḥ's interpretation of the word yešawwe in Hosea 10:1 as "will lie." For Solomon ibn Gabirol qol ha-tor is "the voice of salvation," as in the translation of Targum Jonathan. Moses Ibn Ezra, Samuel ha-Nagid, Al-Ḥarizi and others used the word אֲשִׁישָׁה with the meaning "chalice," an interpretation given to this biblical word by a few commentators. Moreover, the poets could add meanings at their own discretion, as their poetic talents dictated. Since the range of meanings was quite open, the influence of Arabic on poetic language, though on the whole restricted, was felt mainly in the meanings of existing words and not in the creation of new ones.

Influence of Arabic on the Language of Secular Poetry

Whereas the Hebrew of scientific works was deeply influenced by Arabic, the language of poetry was not greatly affected by Arabic other than in the meanings of words and the frequency of rare words. Some words took over the functions of their Arabic cognates but, in addition to the Arabic usage, a source could usually be found for this new meaning in the Bible, sanction was given by the grammarians, and further incentive for the use was provided by the demands of scansion: ל is used to mean "because," as in Arabic, and as found by Ibn Janaḥ in the Bible (Ha-Riqma 55:13), וְאִם to mean "nevertheless," ostensibly like the Arabic wa'in but actually found in the Bible (Num. 36:4; Jer. 5:4).

The process of loan-translation, which so enriched the technical vocabulary of Hebrew in the Middle Ages was restricted in poetic language to literary symbols (in literature these are similar to technical terms): עֹפֶר ("doe"), צְבִי ("deer") for "beloved" – ḡaz̄al in Arabic; זמן ("time"), yamim ("days") for "hostile fate" – dahr, zamān in Arabic; gan ("garden") for "paradise," Arabic ǧanna; perud for "a parting of lovers" – Arabic tafriqa, contrasting with perida, which expressed the separation of death; מוּסָר musar ("right conduct") for "erudition" – Arabic adāb (as in Saadiah Gaon's translation to Prov. 1:2); the expression aḥi musar derives from Arabic, and means "a learned man." Midbar ("desert") for "graveyard," and a few other expressions are reminiscences of similar Hebrew figures of speech to be found in the Bible or the Midrashim. Sometimes a word acquires a new meaning from the range of meanings carried by its Arabic cognate; even in these cases there is usually a biblical source, with Arabic influencing the preference for a particular usage and turning rare expressions into common ones: שָׁב meaning "became" (Ar. ʿāda) – וְשָׁבָה הָאַהֲבָה אֵיבָה ("and love turned to hate"; Al-Ḥarizi). The source of this usage is Isaiah 29:17. עַד for "even" (Ar. ḥattā) הֲבָגְדוּ בִי עֲדֵי אָבִי וְאִמִּי "they betrayed me, even my father and my mother" – Moses Ibn Ezra, with authority for the usage in Judg. 4:16; מַעֲנֶה for "meaning" (Ar. maʿ as in the Targum version of Prov. 1:1); עַם for "people" (Ar. qawm) – as in Judges 9:36 – אהב for "want" (Ar. ahabba).

Words acquire the meaning of their phonetic (and sometimes etymological) counterparts, even when there is basically no identity of meaning. אֲבָל, in addition to its usual meaning of "but," often signifies "and even more," as Ar. bal and authority can be found in Genesis 17:19. The usual synonym for אֲבָל, i.e., אוּלָם also acquired the same meaning: כּוֹכָבָיו צִצִים וְאוּלָם בַּסְּתָיו פָּרָחוּ "his stars were flowers, moreover, in winter they bloomed" (Solomon ibn Gabirol). פֶּלֶךְ is regularly used to mean "the wheel of heaven" (Ar. falak), גִּיל is "generation" (Ar. g ̌il), יעד means "promise" (Ar. waʿda), שָׁם means "afterward" (Ar. thumma). שׁוֹעֵר for Shem Tov *Falaquera means מְשׁוֹרֵר ("poet"; Ar. šaʿir), not the usual medieval interpretations for שׁוֹעֵר in the Bible. יָרֵח בְּהִלּוֹ (Al-Ḥarizi) is "a new moon" – Arabic hilāl. Iggeret haqura (Samuel ha-Nagid) is a "despised letter" – Arabic haqira. Whereas דָּפוֹק means "hasten" in the poetry of Khalfon, in the language of Samuel ha-Nagid and Judah Halevi the root דפק means "flow" (At. dafaqa) and is used as in Arabic figures of speech to describe the flow of tears. חֵשֶׁק is frequently used as love due to its similarity to Arabic ʿašaqa; נַעֲמָה is "an ostrich" (Arab. naʿāma), מְמוּלָּח is "beautiful" (Ar. maliḥ), and רַכָּב אָנִיּוֹת ("traveled on a ship") is also coined after an Arabic expression and הֵנָּה for "here" (Ar. hunā).

In translations the prepositions which follow a verb are much influenced by the source language, especially if it is a spoken tongue (see below on the language of the translations). In the "pure" Hebrew of secular poetry, however, the prescriptions of the grammarians were preserved and the usage of the Bible was followed for most prepositions. Arabic influence explains נָטָה בִּי for "moved me" (Judah Halevi), -נָסַע בְּ for "transported," -עָבַרבְּ "transferred" (Moses Ibn Ezra); in Biblical Hebrew, however, one can find quite similar usages (נִבָּא בִּדְבָרִים ,יָצָא בְּמָחוֹל). Solomon ibn Gabirol wrote וָאֶתְמַהּ מִשְׁלֹמֹה for "surprised at," apparently as in Arabic taʿaǧaba min and was criticized for it by Moses Ibn Ezra (Širat Yisrael 154). There are also characteristic features of poetic language which have no clear links with Arabic: טוּר for "a line of poetry," תֵּבֵל for "the world below," פֶּגֶר for "a body," not a corpse. Features of medieval Hebrew which are common in the language of medieval translations are found only sparingly in poetry: nouns ending in וּת- which were regularly masculine in prose, are treated on a few occasions as masculine in poetry: דְּמוּת נִמְשָׁל (Judah Halevi) הָיָה פְּתַיוּת (Al-Ḥarizi) לְבָבוֹת גְּדוֹלוֹת (Al-Ḥarizi) – this phenomenon is explained below in the section on the translations.

ORIGINAL PROSE WORKS AND TRANSLATIONS

Those poets who disparaged the writers of Arabic as "guarding the vineyards of others" themselves wrote scientific works in Arabic – on philosophy, halakhah, science, poetics, geography, etc. They include Saadiah Gaon, Ibn Gabirol, Judah Halevi, Moses Ibn Ezra and Al-Ḥarizi. Ḥayyuj, Ibn Janaḥ and others even wrote studies of Hebrew grammar in Arabic.

Just as various factors combined to produce a secular poetry in Hebrew, and in biblical Hebrew at that, so the writing of prose and the kind of Hebrew used for translations and original works were interconnected. The Arabs wrote prose in Middle Arabic, with none of the ideal "purity" of the language of the Koran reserved for poetry. And the Jews knew how to write this kind of Arabic (by contrast with the language of the Koran, which required special study). For prose writings they had hardly any linguistic tools to hand, unlike the language of sacred verse which provided a beginning for the writing of secular verse. There was the mishnaic tradition of prose writings on halakhah but Arabic had taken over the function of talmudic Aramaic, and was judged appropriate for writings on halakhah, especially where everyday matters were concerned. Since there was no need to set up a form of Hebrew which should rival the unrhymed Arabic of scientific writings in beauty they could write either in Arabic or in a different Hebrew from that of poetry, with no obligation to observe the rules of "purity" described above. The proportion of mishnaic Hebrew and biblical Hebrew varies from writer to writer. Abraham Ibn Ezra and Abraham b. Ḥiyya (d. c. 1136) both wrote original Hebrew. The former – a grammarian, poet, and biblical commentator – tended to write a biblical Hebrew and preferred forms like לֹא רַק to אַףכִּי ,אֵין…אֶלָּא to כָּלשֶׁכֵּן and so on; the latter used many talmudic expressions: אַגַּב גְּרָרָא ,לְהַלָּן ,סְפֵקָא ,כְּדַאי. The language of prose is mixed, though the writers could write a more biblical Hebrew close to the style of the maqāma, when they chose to, and this applies not only to the poetry of the translators – Abraham Ibn Ezra, Al-Ḥarizi, Ibn Ḥasdai but to passages of rhymed prose interspersed among the testamentary injunctions of Judah ibn *Tibbon to his son or in the letters of Abraham son of Maimonides etc. In his introduction to the translation of Ḥoḇot ha-Leḇaḇot ("Duties of the Hearts") Judah ibn Tibbon explains that he used biblical or mishnaic Hebrew, "whichever seemed closer, and as occurred to me at the time of translation." (It was, for example, convenient for him to translate min ḥaythu by the biblical word בַּאֲשֶׁר which is close to it, but the expression lā illā becomes mishnaic לֹא אֶלָּא.) In its syntax the language of prose is close to mishnaic Hebrew – there are no conversions of tense, no long or short verb forms – but the repertoire of conjunctions is considerably mixed (-לְפִי שֶׁ- ,כֵּיוָן שֶׁ- ,טֶרֶם ,לְמַעַן ,יַעַןאַף עַל פִּי שֶׁ, etc.).

The language of the hundreds of translations carried out from the 11th to the 15th century, and the language of original works written in Hebrew in the style of the translations (e.g., the works of Albo, Crescas, and Levi b. Gershom) is sometimes called Tibbonian Hebrew, after the five generations of Ibn Tibbons who translated into Hebrew innumerable books written in Arabic by Jews and Arabs. Samuel ibn Tibbon called his father Judah "the father of translators," though there had been earlier translations for almost a hundred years. The "translatorese" in original writing derives from the general influence of Arabic, from imitations of language patterns created by the translators, and from the strong attraction of Arabic literature which, though not translated literally, was summarized, with a flavor of the original remaining in the summary. Of course, in kabbalistic literature in original Hebrew there is no more Arabic than sentence patterns derived from the translators, and some terminology, but these traces of Arabic are clearly discernible, both in Kabbalah and in Maimonides' Mishneh Torah.

Though there are nearly as many styles as there are writers or families of writers, it is possible to give a general description of the language of prose that came under Arabic influence. Since it contains many deviations from the forms described by the medieval Hebrew grammarians, Tibbonian Hebrew was sometimes used as a pejorative term for poor, inelegant Hebrew. But it was not usually the result of deliberate carelessness, or lack of respect for the grammarians. Judah ibn Tibbon in the introductions to his translations of Ḥoḇot ha-Leḇaḇsot and Sefer ha-Riqma, and his son Samuel in his introduction to his translation of Maimonides' Moreh Neḇukhim ("Guide of the Perplexed") explained the difficulties arising from the tendency of the translator to adhere closely to the source language text (and they both realized that the similarity of the two languages actually strengthened this tendency). Hebrew was inadequate, they thought, to express the full richness of Arabic, and they asked readers to correct mistakes of language.

From the "explanations of strange words" which Samuel ibn Tibbon appended to his translation of the Guide of the Perplexed, we learn that after he had completed the translation he changed several Arabic-influenced words to better, more Hebrew equivalents: kihun, a borrowing from the Arabic kihāna (though with a Hebrew declension!) was replaced by קסם; the expression עַל דַּעְתִּי modeled on the Arabic ʿalā raʾ, even though it can be given biblical authority (Job 10:7), was replaced by כְּפִי דַעְתִּי, since the greater frequency of the latter expression gives a more Hebrew flavor to the language. In his testamentary injunction to his son Samuel, Judah ibn Tibbon implored him to preserve "the purity (ẓaḥut) of the language," and to beware of Arabisms.

As the language of Spanish poetry should be judged by its appropriateness to prosody and poetic style, so the language of prose should be judged by its suitability to translation and for the skill of the writer as a translator. The principal writers who concerned themselves with the problems of translations were Al-Ḥarizi (Taḥkemoni ch. 18, and his preface to his translation of Maimonides' introduction to the Mishnah), Moses Ibn Ezra (Širat Yisrael p. 112), Judah and Samuel ibn Tibbon in the introductions quoted, and Abraham b. Ḥasdai in his introduction to his translation of Isaac Israeli's Sefer ha-Yesodot. The last three explained that there were "places which were liable to bring the translator into error," and the Ibn Tibbons listed the mistakes that were liable to be caused by too close an adherence to the source language text: confusion of masculine and feminine genders, prepositions, meanings of words, etc. They all described the mistakes as stemming from the "translator's bother." "Translatorese" is evident even in the Maḥberot Iti'el which Al-Ḥarizi translated from the maqāmāt of Al-Ḥariri. Despite their poetic style, with biblical interpolations, Arabic influence is more noticeable in them than in Taḥkemoni, which he wrote in original Hebrew. The latter has also more Arabic usages than his other poetry, since even though not actually a translation it is based on Arab sources.

For a full understanding of the language of prose, the degree of adherence to biblical and mishnaic Hebrew, the scope of Arab influence in all its aspects, and the particular Arabic patterns which affected the Hebrew, it is necessary to classify all the features of the language, distinguishing those which were deliberate innovations (mainly terminology) and those which were accidental, caused by too close an adherence to the source. There are some features which are hardly found outside translations, e.g., a singular verb preceding a plural subject (excluding the verb הָיָה, which was quite commonly found in the singular before a plural). Samuel ibn Tibbon tried to avoid this Arab grammatical rule of concord, as he explains in the introduction to his translation of Moreh Neḇukhim. Features of common occurrence include -ל to denote cause and שָׁב in the sense of "become." The Hebrew language of the beginning of the period, before it became a language of translation, can be judged as a separate entity, capable of influencing those who were faithful to it in their writings (e.g., Abraham Ibn Ezra and Abraham bar Ḥiyya), by contrast with the source language which had all the tools of expression for translation ready to hand. It is worth examining the language of those who did not know Arabic and wrote in the style of the translators, and the language of those who did know Arabic but who did not draw much upon Arabic cultural sources for content, e.g., the kabbalists and writers on halakhah.

It is not surprising that Al-Ḥarizi, who wrote in the first chapter of Taḥkemoni that Hebrew "is narrow but may turn broad to us, short but will suffice for all of us," strove to enrich it from its own sources, avoiding Arab loan-words. But Judah ibn Tibbon, who felt that "Hebrew is insufficient for all purposes of speech" (in the introduction to his translations of Ḥoḇot ha-Leḇaḇot), since biblical and mishnaic Hebrew did not contain the wherewithal for handling new topics, took over many features of Arabic, a richer language in his eyes. However, though it is customary to describe the language of Al-Ḥarizi as simple, correct, elegant, and more biblical (Baneth, Mirsky), it is also full of all kinds of features showing Arabic influence, though not to excess. On the other hand, the Ibn Tibbons also used many specifically Hebrew expressions, out of opposition to Arabic, sometimes consistently and sometimes replacing expressions that showed Arabic influence.

Abraham Ibn Ezra criticized the language of the liturgical poets (in his commentary on Eccles. 5:1) but in the self-same critical passage he wrote that Kallir, as it were, "described the rose on fear," תֵּאַר אֶת הַשּׁוֹשַׁנָּה בְּאֵימָה (as in Ar. waṣafa bi) and that he was "surprised from him" תָּמַהּ מִמֶּנּוּ (Ar. taʿaǧab minhu) and that he fled off the passage," בָּרַח מִן הַפָּסוּק (Ar. haraba min). None of these traces of Arabic are to be found in the poetry of Abraham Ibn Ezra.

Notwithstanding his decision to write a good Hebrew, Judah ibn Tibbon made a rule of preferring to impart the idea with precision rather than "use as good a style as he would prefer" (introduction to Ḥovot ha-Levavot). This method proved its worth; his translation of Ḥobot ha-Lebabot superseded that of Joseph Kimḥi, which was more grammatical but less accurate and his son's translation of the Guide of the Perplexed replaced that of Al-Ḥarizi, which aimed at a greater beauty of language at the expense of accuracy.

The Components of Arabic-Influenced Hebrew
LEXICON

(a) With all the abundance of innovation and wealth of terminology that accompanied the new ideas, the number of words borrowed with their original form and usage is extremely small; most of the borrowings take the form of loan-translations. Most of the terms in philosophic works were translated into Hebrew (one of the Arabic loans is מַשָּׁאִיִּים for "peripatetic," though Al-Ḥarizi translates it הַהוֹלְכִים). The Arabs themselves translated almost all the Latin and Greek terms into Arabic. It was actually in the natural sciences and allied subjects that more words were borrowed. This is immediately apparent if we compare Moses ibn Tibbon's translation of Millot ha-Higgayon ("Words of Logic") where he translated all the technical terms, with his translation of Maimonides' Hanhagat ha-Beri'ut ("Management of Health") where foreign words like אשרוב ("syrup"), names of plants, and foods remain untranslated, as they do in other translations of books on medicine in the Middle Ages. The borrowed words are all in forms and patterns which can easily be adapted into Hebrew and thus absorbed in the language and inflected just like any other word: לַחַן ("melody"), קׁטֶר ("caliber"), קׁטֶב ("pole"), מֶרְכָּז ("center"), אוֹפֶק ("horizon") הָנְדָסָה ("geometry"), תַּאֲרִיך ("history"), עִלָה ("disease"), אַקְלִים ("region"), חׁקֶן ("enema"), עָצָל ("muscle"). Some of them came into general use and became thoroughly Hebraized – אׁפֶק ,תַּאֲרִיךְ – while others were in limited use like עָצָל in the works of Nathan ha-Me'ati ("the Italian Tibbon"), or נוע ("type"), which was rare, and used by Nahum ha-Ma'arabi in the translation of Iggeret Teiman in place of the more usual סוג and מין.

Samuel ibn Tibbon regarded borrowings as the major class of "strange words" and preferred native Hebrew words, changing כִּהוּן to קֶסֶם, and giving to גֶּשֶׁם, in the sense of "body," biblical authority – Isaiah 44:14 – though this is not the usual meaning attributed to this verse in the medieval dictionaries. Al-Ḥarizi who as we have seen wished to widen Hebrew from within, suggested נְקֻדָּה and עַמּוּד instead of מַסְמֵר ,מֶרְכָּז instead of אֲלַכְסוׁן ,קׁטֶב instead of קׁטֶר, and חׁמֶר רִאשׁוֹן instead of הִיּוּלִי. However, in his explanation of foreign words he also includes קׁטֶר and קׁטֶב. He uses two terms for one thing and this suggests that he was unaware of the importance of preserving uniformity of terminology. Other translators also tried to find Hebrew alternatives for loan-words: Nahum ha-Ma'arabi used עַמּוּד in place of קׁטֶב, whilst Abraham Ibn Ezra prefers סַדָּן. Abraham b. Ḥiyya who composed original scientific works before the language of translation had become fixed, quoted Arabic words as such, e.g., "the center of the circle, which in Arabic is called markaz." He uses the term תִּשְׁבּׁרֶת ("geometry," "plane," a loan-translation of Ar. taksīr) and prefers בְּרִיחַ to קׁטֶר, though occasionally he uses קׁטֶב.

(b) Words are sometimes introduced that are similar in sound to their Arabic counterparts, and generally any similarity in meaning or etymological connection is either lacking or very slight. "Grammatical inflection" (in Ar. taṣrīf) is translated by Dunash צֵרוּף, though נְטִיָּה would be more appropriate as a loan-translation (see (c) below).

Judah ibn Tibbon calls apical consonants אוֹתִיּוֹת הַדְּלִיקָה after al-Muḏalaqa, which should etymologically be הַזְלִיקָה. The waw consecutive is called ו״ו עוֹטֶפֶת, from the Arabic ʿaṭafa, מָחוּל means "absurd," like muhāl, פֶּרֶק means "difference" (Ar. farq), גֶדֶר means the mathematical "to the fourth power" (Ar. ǧadr), לָכֵן is used with the meaning "but" (Ar. lākin) and נִצָּב is both "accusative case" and the vowel "a" (Ar. naṣb). Sometimes a biblical word which bears a phonetic resemblance to an Arabic word is used as a translation and the new meaning is given authority by biblical commentary, which was also influenced by the comparison with Arabic. חִידָה is translated as "talk" (Ar. ḥadīth), the translation used by Saadiah Gaon for this word in Proverbs 6:16 and the explanation given by Ibn Janaḥ in his book of roots, under חוד. The Arabic ʿurūq is sometimes translated גִדִּים דּוֹפְקִים, i.e., arteries, and a biblical parallel is found in the word עׁרְקַי (ʿoreqay; Job 30;17), which most medieval commentators interpreted in accordance with the Arabic. It should be pointed out that this kind of innovation is very close to borrowing; the borrowed word, however, is taken over with a change of form to a Hebrew declension, or is attached to an existing Hebrew word.

(c) The most prolific source of word creation was loan-translation. Among the new words created were אֵיכוּת (Ar. kayfiyya), מַהוּת (Ar. mahiyya), כַּמּוּת (Ar. kamiyya), and many other verbal nouns with suffix וּת-, which was used to express abstractions. However, Arabic words were mainly translated by existing Hebrew words. Most of the deliberate innovations used by translators for the enrichment of the means of expression and for accuracy, are in the realm of terminology: בִּנְיָן ("conjugation"; Ar. mabniyya), מִשְׁקָל ("declension"; Ar. wazn), שִׁמּוּשׁ הַלָּשׁוֹן ("language usage") and עֲשִׂיַּת הַלָּשׁוֹן ("language manipulation"; Ar. istiʿmāl), גָּזַר, also כָּרַת ,חָצַב ,קָצַב all meaning "inflect" (Ar. ištaqqa), מִלִּים נִרְדָּפוֹת ("synonyms"; mutarādifāt) מִלִּים מִשְתַּתְּפוֹת ("homonyms"; Ar. muštaraka), מָקוֹר ("infinitive"; Ar. maṣdar), מִקְרֶה ("Abstract noun"; Ar. ḥadaṯ). נָשׂוּא ("predicate" of a verbal sentence; Ar. maḥmul), הַגָּדָה ("predicate" of a nominal sentence; Ar. ḫabar), בְּחִינָה ("aspect"; Ar. iʿtibār), מֻפְשָׁט ("abstract"; Ar. muǧarrad), בְּכׁחַ ("potential") and בְּפׁעַל ("actual"; Ar. bi al-quwwa and bi al-fiʿl;), מַצְפּוּן ("conscience"; Ar. ḍamir), חִבּוּר ("a book"; Ar. taʾlif), הַתְחָלָה ("principle"; Ar. mabdaʾ), הִכָּה ("duplicate"; Ar. ḍaraba). There are also loan-translations which did not provide any technical terminology: בְּשִׁלוֹח ("absolutely"; Ar. biʾiṭlāq), חֻבַּר עַל ("agreed"; Ar. uǧmiʿa ʿalā), - נָאוֹת לְ - ,הִסְכִּים לְ ("matched," "fit"; Ar. wāfaqa) בִּקְצַת הַיָּמִים ("one day," adv., Ar. fi baʿḍi al-ayyām). Since every loan-translation that makes use of an existing word also involves extending the meaning of that word in accordance with the range of meanings of its Arabic counterpart, it is difficult to distinguish between loan-translation and semantic borrowing. Perhaps the fundamental difference between them is the degree of intention. When the motivation is the need to translate an existing Arabic technical term (it is mainly technical vocabulary that is at issue, though non-technical expressions also occur) we speak of loan-translation; when it is the unintended effect of adherence to the Arabic text that leads to certain lexical associations, we speak of semantic borrowing.

(d) The following are examples of extension of meaning by semantic borrowing: עִנְיָן ("meaning"; Ar. maʿ), לָקַח ("begin"; Ar. ʾaḫaḏa), אִגֶּרֶת ("essay"; Ar. risāla), בֵּאוּר ("proof," "lecture"; Ar. bayān), גּוֹבֵר ("common"; Ar. ġālib), רֶמֶז ("advise"; Ar. ʾišāra), דִּין ("religion"; Ar. dīn), כְּמוֹ ("approximately"; Ar. naḥwa), רוֹצֶה ("mean"; as in רוֹצֶה לוֹמַר ("mean to say; Ar. yurīd), בְחׁק ("concerning"; Ar. bi ḥaqq), אֶצְלִי ("in my opinion"; Ar. ʿindi). Sometimes an extension of meaning derives wholly or mainly from a similarity in sound, with or without any etymological connection: זִיֵּף acquires the meaning "deny" from Arabic zayyafa; זִיֵּן ("decorate"), חַג ("pilgrimage"), things which are מְפִיקִים are "suitable" (Ar. muwāfiq).

(e) A feeling for the Arabic language governed the choice of particular Hebrew words, and affected the frequency of words whose use in Hebrew was restricted; this gives a distinctly Arabic flavor to the language. רָאָה ("see") means "think," a use found in mishnaic Hebrew, אָמַר ("say") means "order," חָשַׁק ("love"), חוּשׁ ("feeling") and not רֶגֶשׁ (because of Ar. ḥassa), חִלּוּף ("difference"; Ar. Iḵtilāf) and not שׁנִי or חִדּוּשׁ ,הֶבְדֵּל ("accident") more common than מִקְרֶה or מְאוֹרָע (Arab. ḥadat). What is a borrowing with one writer may be recognized as a legitimate Hebrew usage by another. Samuel ibn Tibbon, for example, quotes from the introduction to Maimonides' Sefer ha-Maddaʿ the expression לֵידַע שֶׁיֵּשׁ שָׁם מָצוּי רִאשׁוֹן to know that there is a God – as clear proof of Arabic usage in the original Hebrew writings of Maimonides, and understandable therefore in a translated text (Introduction to his translation Moreh Neḇukhim – this is a usage of type (b) as analyzed above). Yet since there are rare examples in mishnaic Hebrew of שׁם used to mean "in reality" and not as a locative, it may well be that Maimonides had found this Hebrew source in rabbinic literature for himself. Samuel Ibn Tibbon's father-in-law, Jacob Anatoli, thought that Maimonides found a source for this non-locative use of שָׁם in Ezekiel (in his book Malmad ha-Talmidim p. 113a) as illustrated by S. Abramson, which would make it an example of class (e). Expressions of the form תַּכְלִית הַשְּׁלֵמוּת ("the peak of perfection") a literal translation of Arabic Ġāyat al-Kimāl – were widely used, and a source was found for them in Psalms 139:22 – תַּכְלִית שִׂנְאָה. Typical words include זוּלַת (as a translation for ġayr and duna in their various meanings) and בִּלְתִּי (to translate ġayr and ʿadam). Samuel ibn Tibbon acknowledged that his innovations led to new homonymy when he himself added new meanings to existing words (the fifth class of "strange words").

The following are the new kinds of homonymy created: 1. In addition to its usual meaning in the language, the word received a new technical sense: שֶׁבֶר is a term for the Hebrew vowel ḥireq; and, not particularly technical: ḥida means "talk" (Ar. ḥadaṯ, see above) but also retained the meaning of the ḥidoṯ of the Queen of Sheba, and thus also signifies "allegory" in kabbalistic literature and Maimonides' Mishneh Torah. Sometimes homonymy is transferred from Arabic to Hebrew; bāb in Arabic means "rule," "chapter of a book," and "explanation," and all these meanings were taken over by the Hebrew שַׁעַר. Arabic ḥarf means both "letter of the alphabet" and "particle," and both meanings were transferred to the Hebrew הַעְתֵק .אוֹת translates naqala both in the meaning "translate" and also "hand down (by tradition)," though "tradition" is just as commonly rendered by קַבָּלָה. 2. A Hebrew homonym is paralleled by different Arabic words, in different binyanim or patterns: מַאֲמָר is both "essay" (Ar. maqāla) and "category" (Ar. maqūla); הִשִּׂיג means both "add," a loan-translation of Arabic alḥaqa (fourth binyan), and "appeal," like the Arabic verb lāḥaqa ʿalā (third binyan). 3. The Hebrew homonyms translate two different Arabic words: עוֹבֵר means "possible" – (Ar. ǧaʾiz) and "past" – (Ar. māḏi); מְדַבֵּר is "logical" – (Ar. nāṭiq – "believer" in the philosophy of al-kalām) and also the grammatical term "first person," – (Ar. mutakallim). הִשִּׂיג means "understood" (Ar. adraka) and "added" (Ar. alḥaqa). העתקה is "translation" (Ar. naql), one of the types of metaphor (Ar. maǧāz), and "transmigration of souls" (Ar. tanāsuḥ). 4. The homonymy derives from the falling together of a loan-translation and a phonetic equivalent; גֶּדֶר is both "definition," loan-translation of ḥadd, and "to the fourth power" Arabic ǧaḏr. As regards the principles whereby the vocabulary could be expanded, there are equivalences between the language of scientific prose and the language of Spanish poetry, though the degree of expansion in prose is far greater, due to the needs of writing on new topics rich in new terms and concepts. The case is different with morphology and syntax, where derivations in general and those derived from Arabic in particular abound in prose yet are hardly found at all in poetry. Since the language of prose was not subject to the principles of ẓaḥut ("purity") which mainly affect the formal aspects of grammar, and since its counterpart was an intermediate variety of Arabic, which had a different degree of adherence to the strict rules of classical Arabic, it even deviated from the rules of grammar established in the grammatical writings of the period. It was not an elegant language, and its foreign features were conspicuous, but its freedom to innovate helped to fashion it into a precise language of scholarship, capable of expressing abstract, scientific ideas.

The following are the salient features of the language of prose:

NOUN MORPHOLOGY

The use of the suffix morpheme ־ִי (called in Ar. nisba) to turn a noun into an adjective meaning "possessing, related to, having the quality of" was productive, almost automatic: גַּשְׁמִי ,שִׁמּוּשִׁי ,דִּבְּרִי; Arabic gave new life to this suffix (found in the Bible), which most frequently occurred as in Arabic without the infixed nun after short words or words ending in a vowel: גּוּפִי ,חוּשִׁי ,צוּרִי ,תּוֹרִי ,רוּחִי (for צוּרָנִי ,גּוּפָנִי, etc.). This morpheme does not feature in the linguistic innovations of Saadiah Gaon and was rare in poetry, which used participles instead: מְדַבֵּר for דִבְּרִי ("logical"), or phrases like בַּעַל גֶּשֶׁם for גַּשְׁמִי and אֲחִי מוּסָר for מוּסָרִי etc. Moses Ibn Ezra also considered this kind of innovation contrary to analogy (unproductive) for the morphology of "pure" poetic language (Širat Yisrael, 151, apropos the derivation of פְּנִינִיָּה from פְּנִינָה, which is "mere cleverness").

In varieties of Hebrew which were closer to mishnaic Hebrew, and more restrained in their enthusiasm for Arabic, this derivation was replaced by the typically mishnaic pattern: קַטְלָן e.g., in the writings of Abraham bar Ḥiyya – שַׁקְרָן for שִׁקְרִי ("something containing lies" not "one who lies"), עַמְלָן for מַעֲשִׂי, "practical" (contrasted with שִׂכְלִי "intellectual"), דַּבְּרָן for דִּבְּרִי ("logical"). The morpheme-uṯ creates abstract nouns and can be combined with nouns: נֶצַח נִצְחוּת"eternity," רֵיקוּת – רֵיק"vacuum," רְגִילוּת – רָגִיל "regularity," גַּשְׁמִיּוּת – גַּשְׁמִי "corporeality," and with verbal nouns: הֵאָצְלוּת ,הִפָּעֲלוּת ,הִתְרַשְּׁמוּת. Sparing use had been made of this device by the liturgical poets (דַּכְאוּת ,וָתִיקוּת), apparently under the influence of the infinitive with the suffixed definite article in Aramaic (אִתְעָרוּתָא), but in the language of Saadiah Gaon it is not used at all and it is extremely rare in Spanish poetry. In prose it became indispensable, one of the preferred productive morphemes, even in the prose writings of the poets (Abraham Ibn Ezra uses דַּגְּשׁוּת and דַּיְּקוּת). In Arabic the suffixed morpheme-uṯ was borrowed from Aramaic, and is quite limited in distribution; this is not an example of Arabic influence.

The use of the masculine gender for nouns formed with וּת- is characteristic, and though Samuel ibn Tibbon found authority for this in the Bible (אָחַז בְּשָׂרִי פַּלָּצוּת "a shudder seized me" in the introduction to his translation of The Guide for the Perplexed, quoting Job 21:6) the abundance of these forms in the masculine clearly derives from Arabic: Nouns ending in וּת-are masculine in Arabic, and a considerable number of the new words formed in Hebrew with this morpheme derive from Arabic infinitives which are also masculine: גַּשְׁמוּת (Ar. taǧsim̄), נַצְחוּת (Ar. taʾbīd), הִשְׁתַּדְּלוּת (Ar. iǧtihād), and many more. Use of the masculine gender also spread to those words, few in number, where the suffix corresponds with the feminine abstract noun suffix in Arabic: אֵיכוּת כַּמּוּת ,מַהוּת, though these words were also used in the feminine. Words with the suffix ית- were also used in the masculine and Samuel ibn Tibbon found authority for this usage also in the Bible (וְאַחֲרִיתְךָ יִשְׂגֵּא; Job 8:7). Sometimes words are used in the grammatical gender of their Arabic counterparts: אֱמֶת is masculine – אֱמֶת גָּמוּר (Arab. ḥaqq); דַּעַת is also masculine דַּעַת בָּרוּר (Arab. 'alm). Words used in the feminine include אִי ,טֶבַע (Ar. ǧazīra), כּׁחַ (Ar. quwwa), מִנְהָג (Ar. ʿāda), כַּדּוּר (Ar. kurra), and סִפּוּר (Ar. qiṣṣa). They are of course also used in the masculine, in accordance with the tradition of the language, and the proportion of Arabic or Hebrew usages varies with the writer's talent and grammatical knowledge. In addition to the masculine use of אֱמֶת the form אֲמִתָּה is very common, due to Arabic ḥaqiqah. When a masculine plural is formed by adding -oṯ, attributive adjectives characteristically take the same ending: סְפֵקוֹת גְּדוֹלוֹת ,מְקוֹמוֹת יְדוּעוֹת ,סוֹדוֹת עֲצוּמוֹת, (but שֵׁמוֹת מִשְׁתַּתְּפִים). This formal correspondence may have been helped by the rule in Arabic that inanimate plurals take adjectives in the feminine singular. When the translator was faced with asrār ʿaṣima, the adjective was drawn towards the feminine, though not feminine singular; since there was no precedent in Hebrew for a structure like "סוֹדוֹת גְּדוֹלָה" such a form was naturally rejected. Middle Arabic had also begun to challenge the rules of congruence in Classical Arabic, and tended toward greater uniformity. Since Arabic had only one form and syntactic usage for what are usually described nowadays as the infinitive (קָטוֹל) and the verbal noun (קְטִילָה), both these forms are used interchangeably in the language of prose, e.g., הַרְחָקַת הַגַּשְׁמוּת וְהַעֲמִיד הָאַחְדוּת "removal of corporeality and establishment of uniformity." Arabic Ġāyat al-Taḥaffuẓ is translated both תַּכְלִית הַשְׁמוֹר and תַּכְלִית הַשְּׁמִירָה and this is the reason that the use of the infinitive with the definite article is common: הַהֵעַשׂוֹת ,הַהִתְעַצֵּל (Saadiah Gaon also makes no distinction between these forms. The inability of the infinitive to take the definite article begins with Samuel ha-Nagid, and in the language of poetry neither the absolute nor the construct infinitive is used as verbal nouns.) The abundant use of fused construct forms (i.e., without šel) can be attributed to the influence of Arabic construct forms, as can the use of the definite article before an adjective in the construct form: הַחֲזק הַלֹבֶן ,הַמְחֻיַּב הַמְּצִיאוּת and especially before the comparative הַיּוֹתֵר חָשׁוּב. This is the Arabic feature of marking as definite any construct form which is not a noun, the so-called "unreal construct." The tendency to use two construct forms with a single dependent noun – פּוֹשְׁעֵי וֶרִשְׁעֵי יִשְׂרָאֵל ("the evil and wicked of Israel") – may be due to the influence of Middle Arabic.

VERB

In the introduction to his translation of Guide of the Perplexed, Samuel ibn Tibbon acknowledges the tendency of a translator to be drawn towards the קֶשֶׁר (his term for preposition) used with the verb in Arabic, e.g., qibbel le- instead of qibbel min. Arabic prepositional usages, infrequent in the language of poetry, are quite common in prose:

מִתְפַּלֵּא מִן ,נִפְלָא מִן (Ar. taʿaǧaba min); מתרעם מִן (Ar. ġaḍaba min); חקר עַל (Ar. fataša ʿan); מקיף ב (Ar. ʾaḥāta bi), הבטיח ב (waʿada bi); חסר אל (Ar. iftaqara ilā); and so are התנגד עַל ,קרוב מִן ,גינה מִן ,לא יתכן מִן, etc.

Sometimes the preposition adds a specific meaning as in Arabic: אָמַר בְּ- = הֶאֱמִין for "believe in" (Ar. qāla bi). Intransitive verbs of motion are made transitive by the "causative bet" like in Arabic; this usage is also found in poetry, but to a much greater extent in prose; -הִתְגֵּלְגֵּל בְּ ("rolled"). -עָף בְּ "set flying"), etc.

Judah and Samuel ibn Tibbon were aware of the influence of the Arabic binyanim on the Hebrew verb, but apparently did not consider this such a serious defect as the influence of the prepositions; in poetry too the amount of analogical formation and innovation in the verb was greater than in the noun and the particles. The main development was the increase in the use of the hitpaʿel, which translated three Arabic binyanim: tafāʿala (as הִתְחַסֵּר, Ar. tanāqasa); tafaʿʿla (פּׁעַל מִתְעַבֵּר "transitive verb," Ar. mutaʿadd, הִזְדַּכֵּר "remember," Ar. tazakkara, הִסְתַּפֵּק "be in doubt," Ar. tašakkaka); iftaʿala (שֵׁמוֹת מִשְׁתַּתְּפִים "homonyms," Ar. muštaraka, מִתְתַלְּפִים "different," Ar. muḫtalifa, הִסְתַּכֵּם "agree," Ar. ittafaqa, מִתְאַחֵר "later, following," e.g. הַגְּאוֹנִים הַמִּתְאַחֲרִים, Ar. mutaʾaḫḫir).

In Middle Arabic, these binyanim had largely supplanted the "internal passives," hence: הִתְבָּאֵר (Ar. tabayyana) instead of בּׁאַר (Ar. buyyina), הִתְיַלֵּד "was created" (said of rainbow, water, etc., Ar. tawallada), etc.

The use of hitpaʿel as a passive in place of passives with internal vowel modification is not simply continuation of mishnaic practice, since the Ibn Tibbons also introduced many forms of "internal" passives. The increase in the use of passive conjugations, in prose and in poetry, is attributed to the influence of Arabic. The "internal" passives – puʿal, and nipʾal – were used mostly in impersonal structures: יְעֻיַּן ,יְזֻיַּן ,יֻכְסַף ,יֵרָצֶה יְחֻיַּב ,יְסֻפַּק. New auxiliary verbs were created חָזַר ,שָׁב in the sense of "become" (Ar. ʿāda, raǧaʿa), לֹא סָר ,לֹא זָז meaning "keep (doing something)" (Ar. mā zāla) and עָשָׂה ,שָׂם, "make," with an objective complement – עָשָׂה הַסָּפֵק בָּרוּר ("make doubt clear, clarify doubt"; Ar. ǧaʿala). Tenses converted by waw were almost no longer employed. The modal forms of the verb, jussive and cohortative, were scarcely used in medieval Hebrew prose. In poetry the long and short forms of the imperfect were used simply as morphological variants, not expressing modal ways as they were understood by Hebrew grammarians to appear in the Bible, without any influence from verbal patterns found in Classical Arabic. The language of prose dispensed with these forms, since it had no need of a multiplicity of forms for embellishment. Middle Arabic, which had lost some verb forms, may also have contributed to the general picture and helped to eliminate long imperfects in וּן since in North Africa and Spain it was the short forms which were used in the plural. A Past Continuous or a Past Habitual, like the Arabic kāna yaf ʿal occurs: hava yabo (a usage also found in Saadiah Gaon and in the maqāmāt of Al-Ḥarizi). The combination of כְּבָר and the imperfect is used to express possibility, by analogy with Arabic qad; in fact, כְּבָר as an equivalent for qad is increasingly used, both in the language and in grammatical description. Ibn Janaḥ (Sefer ha-Šorashim, s.v. כְּבָר) explained that כְּבָר was like Arabic qad, and expressed "the existence of a thing." By using כְּבָר as an equivalent of qad the following tenses were formed: pluperfect: הָיָה כְּבָר עָשָׂה; future perfect: יִהְיֶה כְּבָר עָשָׂה. This pattern is found mainly in the language of translation, but also occurs elsewhere (e.g., in the writings of Crescas, 14th century). However, it had already begun to fade from the grammatical stock of medieval Hebrew a few generations after the end of the period of the translations. But the use of כְּבָר plus the perfect to signify time and for emphasis כְּבָר יָדַעְתָּ ,כְּבָר נִתְבָּאַר equivalent to the Arabic qad, was much more in line with the spirit of the language, and Arabic accounts merely for its widespread distribution. As in Arabic כְּבָר was placed before the verb, as in mishnaic Hebrew where this usage originates, and not after the verb, as occurred occasionally in mishnaic Hebrew and quite commonly in varieties of Hebrew influenced by languages where the equivalent of כְּבָר (schon, déjà) come after the verb. There are, however, examples of כְּבָר occurring after a verb, mainly when the verb is in a subordinate clause: מַה שֶּׁשָּׁמַעְתִּי כְּבָר מִטַּעֲנוֹת הֶחָבֵר ("… those statements of the companion which I have already heard"; beginning of the Kuzari).

SYNTAX

The use of the demonstrative without he (the definite article) before a noun with he is very common: זֶה הָאִישׁ ,אֵלּוּ הַדְּבָרִים. The structure with the demonstrative before the noun and no he – אוֹתוֹ הָאִישׁ, as in mishnaic Hebrew – is preferred to הָאִישׁ הָהוּא, the biblical form. Though there are a few examples in the Bible to serve as precedents (זֶה הַיּוֹם Ps. 105), and an equivalent structure in late Aramaic (הָדֵין עָלְמָא), Arabic was certainly the major factor: hādhā al-walad, dhalika al-walad. This usage is not found at all in the writings of Saadiah Gaon and hardly in poetry, but occurs frequently in the original prose writings of Abraham b. Ḥiyya, Maimonides, and Al-Ḥarizi. However, there are many places where the Ibn Tibbons used forms like הַדְּבָרִים הָהֵם ,הָאִישׁ הַהוּא in their translations, even when this meant deviating from the word order that confronted them in the Arabic text: ḍalika al-raǧul. The use of relative clauses with no conjunction after an indefinite antecedent (-אִישׁ עִיֵּן בְּ for -מְלָאכָה צְרִיכָה נִסָּיוֹן ,אִישׁ שֶׁעִיֵּן בְּ for מְלָאכָה שֶׁצְּרִיכָה נִסָּיוֹן) is also the exact counterpart of an Arabic structure (sifa). This structure can also be given biblical authority (שָׂרִים זָהָב לָהֶם) but comes in prose much more than in poetry. Elegant translators added the definite article, in places where it did not occur in Arabic, in order to bring the structure closer to the form prevalent in mishnaic Hebrew and to a considerable extent also in biblical Hebrew (compare Aḥiṭub's translation of Millot ha-Higgayon with Moses ibn Tibbon's). Moreover, the Ibn Tibbons added quite a few relative clause markers (הַ- ,אֲשֶׁר ,שֶׁ) where none existed in Arabic, and translators like Al-Ḥarizi and Aḥiṭub are not free from asyndetic relative clauses. Such clauses are also found in original texts which were influenced by the language of the translations, e.g., Beit ha-Beḥira by Ha-Me'iri, written in Provence at the end of the 13th century.

Relative clauses were also formed, on the modal of the Arabic naʿt sababī, in which the adjective or participle is predicative to a following noun, and agrees with it in number and gender, but preserves an indirect link with the antecedent with which it shares the same category of deixis, definite or indefinite: הַמִּדּוֹת הַמְסֻפָּר בָּהֶן הַבּוֹרֵא ("the qualities (feminine plural) attributed (masculine singular) to the Creator (masculine singular)"). This structure occurs most frequently when the predicate in the relative clause is a passive participle and impersonal: צׁרֶךְ הַלָּשׁוֹן הַמָּעְתָּק אֵלָיו (Judah ibn Tibbon's introduction to Ḥoḇot ha-Leḇaḇot) הַדְּבָרִים הַמֻּזְהָר מֵהֶם וְהַמְצֻוֶּה בָּהֶם ("things (masculine plural) warned against (masculine singular) and commanded (masculine singular)" (Moses ibn Tibbon's translation of Sefer ha-Miẓwot). Also based on Arabic is the common structure with מִן…מִי ,מִן…מַה (and similar structures with other words replacing מַה and מִן), where the first part of the sentence functions restrictively: מִן הַטְּעָנוֹת וְהַתְּשׁוּבוֹת מַה שֶּׁיֵּשׁ אִתִּי (the beginning of Ibn Tibbon's translation of the Kuzari; "what I have of claims and answers," i.e., those claims and answers that I have). Similarly מֵעָוֶל וְחָמָס ,מַה שֶּׁהִתְאַמֵּת וְהִתְבָּרֵר מִמַּאַמְרֵי הַתּוֹרָה מַה שֶּׁעָשׂוּ לָנוּ הַמִּצְרִים, etc. This structure, modeled on Arabic, survived at least until the 18th century. Gershon b. Solomon composed his Šaʿar ha-Šamayim in the second half of the 13th century. All scholars are agreed that he did not know Arabic. His book, "a breviary of the wisdom of nature," was based on scientific works translated from Arabic. He picked up this structure, new to Hebrew, from the books he studied and he understood how to use it correctly. He writes for example: חוּשׁ הָרְאוּת לְכַוֵּן הַנָּאוֹת לוֹ מִמַּאֲכָל וּמַשְׁקֶה ("the appropriate from food and drink," i.e., those foods and drinks which are appropriate). In the spirit of Arabic are the many object-noun clauses in place of infinitives: נִרְצֶה שֶׁנֵדַע ("we shall want that we shall know"; instead of נִרְצֶה לָדַעַת "we shall want to know"; compare the Arabic nurīdu an naʿrifa; נִצְטַוֵּינוּ שֶׁנַּעֲנִישֵׁהוּ ("we were ordered that we should punish him") instead of נִצְטַוֵּינוּ לְהַעֲנִישׁ אוֹתוֹ ("we were ordered to punish him"), רָאִיתִי גַּם כֵּן שֶׁאֲחַבֵּר חִבּוּר ("I decided to compose….") Subordinate clauses are also common after words such as רָאוּי ,צָרִיךְ, etc. and are modeled on Arabic ʾan clauses introduced after such verbs e.g., צָרִיךְ שֶׁנִּתְבּוֹנֵן ("it is necessary that we look") for אָנוּ צְרִיכִים לְהִתְבּוֹנֵן ("we must look"; compare the Ar. yaǧibu ʾan).

By contrast, infinitives are frequent in place of subordinate adverbial clauses of time, purpose, reason, and comparison: אַחֲרֵי בָאֲרִי for "after I had explained" (Ar. baʿad tabayyuni); לִהְיוֹתוֹ for "because he is" (Ar. likawnihi); כְּאָמְרוֹ ("as he says" or "as it says in the text"; Ar. kaqawlihi). Under the influence of Arabic the use of אֶת before a direct object diminished and the use of cognate objects increased e.g., תָּמַהּ תְּמִיהָה. The circumstantial use of participles is common אָמַר הַנָּבִיא מִתְחַנֵּן ("the prophet said, imploring"; Ar. qāla al nabī mutašafiʿan). The use of the objective complement is also frequent: יָשִׂים הַסָּפֵק בָּרוּר ("he will make the uncertainty clear," i.e., will clarify it.).

Literal translation produces structures which are the exact image of the original Arabic text: וּתְחִלַּת מִי שֶׁמָּצָא זֶה הַדַּעַת i.e., the first who found… (Ar. awwal man ʿamada ʿalā). The use of the prefix מִי – for listing details and explanations (called in Ar. the mīm al-mubayyina) was transferred from Arabic, when it was not translated by such words as כְּגוֹן ,כְּמוֹ ("such as") which is the usual method. Though not mentioned by grammarians like Ibn Jannaḥ, it is found in untranslated Hebrew, even in thoroughly Hebrew contexts such as Maimonides' Mishneh Torah: כָּל הַבְּרוּאִים מִמַּלְאָךְ וְגַלְגַּל ("all created things, such as, angels…" Yesodei Torah 4, 1).

In the Bible the word אָמְנָם adds emphasis and by virtue of this usage is employed to translate typically Arabic structures. Due to its phonetic similarity to Arabic ammā it is used to emphasize the subject: וְאָמְנָם הָרִאשׁוֹנוֹת צְרִיכוֹת בֵּאוּר ("as for the first ones, they need proof"). And because of its phonetic similarity to Arabic innamā it can emphasize a following predicate. Other words are also used to translate ammā and innamā – -אֲבָל ,אִם ,אוּלָם ,אֲשֶׁר לְ – even by the Ibn Tibbons. Al-Ḥarizi, who to some extent preferred אֲבָל, very often tried to side-step such structures altogether. However, the use of אָמְנָם to emphasize what follows, a structure modeled on Arabic syntax, is very common in works which drew their Arabic inspiration from the translations written in the 13th and 14th centuries: they occur frequently, for example, in the writings of Cordovero, who lived in the 16th century.

Translators who were nearly always led to render the Arabic lā ilā by לֹא…אֶלָּא very seldom managed to use לֹא אֶלָּא to translate the Arabic innamā. The merits of original, untranslated texts are noteworthy by comparison with the language of the translations. Samuel ibn Tibbon translated the words of Maimonides (Guide of the Perplexed 2, 44): בַּמַּרְאָה אוֹ בַחֲלוֹם הַנְּבוּאָה אָמְנָם תִּהְיֶה ("prophecy will only be in a vision or a dream"; in Ar. the word innamā comes after the word for prophecy). But Maimonides himself wrote (Hilkhot Yesodei Torah 7, 2) אֵין רוֹאִין מַרְאָה בִּנְבוּאָה אֶלָּא בַחֲלוֹם ("no vision of prophecy is seen except in dream…").

MAIMONIDES

This example from Maimonides of a Hebraic structure, …אֶלָּא לֹא rather than the use of אָמְנָם in imitation of an Arabic pattern, is not unique. Asyndetic relative clauses are rare in the Mishneh Torah, and many of the laws begin after the fashion of mishnaic Hebrew: צוּרוֹת שֶׁעֲשָׂאוּם ("shapes that have beenmade"), אֶפְרׁחִים שֶׁקַּנְּנָה בָהֶם ("chicks for whom a nest has been made"). נָבִיא שֶיַּעֲמׁד לָנוּ ("a prophet who shall represent us") etc. Subordinate clauses are preferred to infinitives, and particularly worthy of comparison are the many occasions where Maimonides uses כְּמוֹ שֶׁנֶּאֶמַר or בְּעִנְיָן שֶׁנֶּאֶמַר as against כְּאָמְרוֹ, which Moses ibn Tibbon uses in his translation of Maimonides' Sefer ha-Miẓwot.

In vocabulary also, the Hebrew of Maimonides tends to be free from Arabic influence. Whereas Samuel ibn Tibbon writes of קׁטֶב הַתּוֹרָה ("the pole of the Law"; in his translation of Guide of the Perplexed), in line with the Arabic figure of speech, Maimonides himself wrote עַמּוּד הַחָכְמוֹת – "the pillar of wisdom" (introduction to Sefer ha-Maddʿa) – see above on the respective uses of עַמּוּד and קׁטֶב. He wrote גּׁלֶם וְצוּרָה and not גֶּשֶׁם וְצוּרָה ("matter and form"), נִתְבָּרַר and not נִתְבָּאַר (explained), which is a loan translation of tabayyana, מְאׁרָע and not חִדּוּשׁ (event; Ar. ḥadaṯ) though he did also use the words נִתְבָּאַר (meaning "explained and proved") and יִתְחַדֵּשׁ (meaning "take place"). Maimonides took great care with the language of Mishneh Torah; he wanted it to be "clear and precise" (as he wrote in the introduction) and chose to use not the language of prophecy or the language of the Talmud (i.e., Aramaic) but "the language of the Mishnah so that it will be easy for the majority" (as he wrote concerning the Mishneh Torah in his introduction to Sefer ha-Miẓwot). In fact, he used typically mishnaic forms more than was usual in the language of the translations: הֵיאָךּ ,לֵילָךּ ,לֵידַע – and particularly the use of a proleptic pronoun נִתְּנוּ לוֹ לְמשֶׁה, which is not a feature of Arabic-influenced medieval Hebrew.

For all his conscious preference for "the language of the Mishnah," Maimonides interlarded his prose with many biblical expressions, not just vocabulary items but whole phrases in a rhetorical style replete with biblical quotations: מָרֵי נֶפֶשׁ ("bitter of soul," Prov. 31:6), מְפֻזָּר וּמְפׁרָד ("dispersed and scattered," Esth. 3:8) בְּלוֹיֵי סְחָבוֹת ("cast off remnants," Jer. 38:11), כְּמַר מִדְּלִי ("a drop in the bucket," Isa. 40:15), etc. There are undoubtedly traces of Arabic influence in the language of Maimonides, but they nearly all derive from Arabic features in Hebrew texts written by his predecessors, and most of them have their roots in Hebrew: prefixed ל – to indicate cause, prefixed מ – for exemplification (see above), עִנְיָנוֹת רְחוֹקוֹת instead of דַּעַת ;עִנְיָנִים רְחוֹקִים in the masculine, אֲמִתָּה for מִסְתַּפֵּק מִן ;אֱמֶת meaning "be doubtful about…"; הָעוֹבְדֵי אֱלִילִים ("idol worshippers") with the definite article preceding the construct form; הִשִּׂיג meaning "understand"; הֶעְתִּיק meaning "hand down by tradition" (Ar. naqala); the technical terms עִקָּר "principle," צוּרָה ("form"); עִלָּה וְעָלוּל ("cause and effect"); מָצוּי רִאשׁוֹן ("first entity" – a term for God). A typical feature of his prose is the translation into Hebrew of most of the Aramaic expressions in the Talmud: שָׁלַם הָעִנְיָן for סְלִיק עִנְיָנָא ("the matter is complete"); כָּל הַמִּתְגַּבֵּר זָכָה for כָּל דְּאַלִּים גְבַר ("whoever is in power wins");בֶּן מֶצֶר for בַּר מִצְרָא (an immediate neighbor); and many more. Maimonides' language is closest to the style of learned medieval Hebrew in passages of philosophical reflection (especially in Yad, Yesodei ha-Torah), and it is these sections which best show how superior and "Hebraic" his style is by comparison with the language of the Ibn Tibbons.

ITALY

In the year 1054, in Italy, Ahimaaz b. Paltiel, descendant of a line of liturgical poets, wrote his Megillat Aḥimaʿaẓ, the genealogical record of his family. The language of his book is naturally akin to that of the liturgical poets. There are many features of mishnaic Hebrew (binyan nitpaʿal, proleptic use of pronouns as in יֶשׁ לוֹ לְאָדָם, demonstratives as in אוֹתָם הַיָּמִים, words such as שֶׁכֵּן ,צָרִיךְ ,סָבוּר etc.), enriched by biblical Hebrew in the form of interlarded quotations, morphology, syntax and vocabulary: waw conversive, infinitival phrases as in בְּבוֹא ,בְּהַגִּיעַ, lack of innovation in puʿal and so on. A form like לֵילְכָה, with a mišnaic infinitive – לֵילֵךְ – and a biblical lengthening, testifies to the blend of language varieties! Like the liturgical poets that preceded him, he combined -כְּ with a perfect tense – כְּרָאוּ ,כְּבָט, made much use of allusive phrases (שׁוֹשַׁנָּה, "rose" = Israel), and employed the typically liturgical vocabulary: רָשָׁה (for "said, command"), בְּכֵן ,גֶּשֶׁם ,פְּצִיחָה (for "then") etc. Traces of Italian influence on the language are slight but well defined (פֻּרְקוֹן for "fork," and see below with respect to the use of טֶרֶם). The initial use of a demonstrative (זוֹ הַמְּדִינָה) in the language of Aḥimaʿaẓ may derive from late Aramaic, or from the occasional use of the structure in mishnaic Hebrew, or possibly even from Italian or Arabic, as in the case of מִזֶּה הָעִנְיָן in the writings of Shabbetai Donnolo, who preceded him by almost 100 years. This book was written before the language of Spanish poetry had taken definite shape, but Immanuel of Rome, who lived in the middle of the 13th century, and who wrote a clearly biblical Hebrew, studded with quotations, though enriched with mishnaic features and even references to the Mishnah, was well versed in Spanish poetry, as he himself bears witness. Since he followed the trend of Spanish poetry and drew upon its language as a source of inspiration, he makes use of innovations and words whose frequency has risen under the influence of Arabic: סָר meaning "stop," שָׁב for "become," אִנּוּן for "grief" (Ar. ʾanna), כְּמוֹ for "approximately," תִּשְׁבּׁרֶת ("geometry"), גֶּשֶׁם ("body"), אֱלוֹהִי ("theological"), לִמּוּדִי ("mathematical"), the pattern מִן…מַה (see above, p. 1631f.) and so on. The imprint of Italian is inestimably greater on his style than on that of Ahimaaz; he wrote poems in Italian and introduced the sonnet into Hebrew poetry. Italian accounts for the use of the following words in the masculine: אֶצְבַּע תֵּבֵל ,בֶּטֶן ,רֶגֶל ,צִפּוֹר, and of the following in the feminine: זָקָן ,שַׁעַר ,חֶסֶד, etc. פֶּרַח, a name of a coin, is a loan translation of the Italian "florin" and the pattern ׁהַשֶּׁלּו is a reflection of il suo.

Shabbetai Donnolo, a physician who lived in the 10th century in southern Italy, also wrote in a style blended of mishnaic and biblical Hebrew, and the result is entirely different from the language of the Ibn Tibbons. In addition to technical terms borrowed from Latin and Greek, his language also shows Italian influence: בֶּטֶן in the masculine (like il ventre), טֶרֶם and מִטֶּרֶם take on the meanings of Italian avanti i.e., first of all, better: נָאֶה לָרוֹפְאִים לֵידַע טֶרֶם הַבְּשָׂמִים ("it is good for doctors, first of all to know perfumes"). Ahimaaz too uses terem in the sense of "before anything else."

Obadiah of Bartinoro wrote his letters from Ereẓ Israel (in the middle of the 15th century) in a language basically biblical but enriched with mishnaic features, and showing signs of Italian influence, apart from a few usages, mainly in loan words – cottimo for "piece work," capitano for "captain" etc.

THE KARAITES

The Hebrew of the Karaites has not yet been described as a distinct variety. For the moment it must suffice to say that the rhymed polemical writings resemble those of Saadiah Gaon and the language used by the ge'onim in their liturgical poetry. The writings of Daniel al-Qūmisī' (ninth century), Sahl b. Maẓli'aḥ, and Solomon b. Jeroham (10th century), for example, are largely biblical in style, and richly studded with quotations, but also contain freely derived verb forms, in all the binyanim, and noun declensions of which the most productive are קְטִילָה ,קִטלוֹן and קֶטֶל. The Karaites fought against the oral tradition, and Saadiah Gaon countered their arguments by pointing out the indispensability of mishnaic Hebrew for understanding the Bible (in his Perush Shivʿim Millim); their language, however, is not a pure biblical Hebrew. None of them abstained completely from mishnaic usage, not merely as regards such content words as were vital in the debate on oral law (הֶתֵּר ,אִסּוּר ,גְּזֵרָה ,תַּקָּנָה ,מִדְרָשׁ, etc.) but also structures and form words characteristic of the Mishnah: binyan nitpaʿal, צָרִיךְשֶׁ- ,רָאוּי לְ- ,אוֹתָהּ הַשָּׁנָה, etc. And their vocabulary included words typical of liturgical poetry: רָשָׁה ("say, command"), צָרַח ,פְּצִיחָה, etc.

The Karaites were much influenced by Arabic culture, and their prose style is therefore marked by the influence of Arabic; it is very close to the language of the Ibn Tibbons (see Eškol ha-Kofer by Hadassi, written in Istanbul in the 12th century). But its specific features are worth special study. There are certain terms characteristic of Karaite Hebrew, some of which occur nowhere else: נִזְכָּרִי and נְקֵבָתִי ("masculine" and "feminine"), עַדְפָנוּת ("advantage"), כֶּתֶם ("impression"), אֶפַע ("event"), הֱיוֹתוּת ("existence"), and so on.

For recent studies of Karaite Hebrew see Maman in Bibliography.

Samaritans

From the 13th and 14th centuries onwards the Samaritans composed prayers and other works in Hebrew influenced by Arabic. A full description has not yet been made; Cowley has offered initial research as has Z. Ben Ḥayyim (Tarbiz 10). A necessary line of investigation will have to be how this Hebrew could exist independent of extra-biblical Hebrew (Ben Ḥayyim, Lešonenu Laʿam, 1969).

See also Ben Hayyim, Ivrit va-Aramit Nusaḥ Shomron, vols. 1–5 (1957–77).

PROVENCE AND NORTHERN FRANCE

The Jews of Provence came under the influence of Spanish Jewry; it was in the towns of Provence (mainly Narbonne and Lunel) that the work of the Ibn Tibbons and the Kimḥis in translation was carried out. It is not surprising, therefore, that Menahem ha-Me'iri from Provence (14th century) wrote his book Beit ha-Beḥira in a style containing all the typical features of "Tibbonian" Hebrew. The influence of Arabic is marked, especially in the introduction, and not in the body of the work, which is a summary of halakhic judgments. Nevertheless, even the actual discussion of halakhah shows far more Arabic influence than the Hebrew of Maimonides' Mishneh Torah.

Unlike the Jews of Provence, the Jews of northern and eastern France had strong cultural and social ties with the Jews of western Germany – "Ashkenaz" – and it was in the towns of Champagne, the Rhine Valley and Lorraine that Ashkenazi Hebrew was fashioned. The Hebrew of Rashi and the French authors of Tosafot (talmudic commentary) is close to Ashkenazi Hebrew, as regards its sources and constituent elements, but the language of influence is Old French, not Middle High German; the influence is much smaller than the corresponding influence of German on Ashkenazi Hebrew.

The influence of French accounts for the increased use of the verb ʿasa in the Hebrew of Rashi, by analogy with "faire,"e.g., עָשָׂה בְּרָכָה, though such forms do occur in mishnaic Hebrew. Also from French is the use of בַּיִת and בָּשָׂר in the feminine. But the French background is most marked in the direct quotation of Old French words in order to explain the Hebrew: "רׁשֶׁם, in French cogneau," etc. The language of Rashi is generally excellent, accurate Hebrew; it is largely mishnaic, enriched with biblical words and forms, even his commentary on the Talmud, e.g., תִּפְעַלְנָה, infinitival clauses like בְּבוֹא ,בְּשׁוּבִי, the lengthened imperfect, and figurative expressions such as לִבּוֹ דָוֶה עָלָיו ("his heart grew faint" – see Lam. 5:17), הוֹן עָתָק מְאׁד ("very great riches" – see Prov. 8:18), though on a modest scale compared with the rhetorical figures in the rabbinical style. In his commentary on the Bible he made sparing use of Aramaic, and only in fixed expressions: כְּמָה ,דְאִיתְמַר דִכְתִיב, and a wide use of the Aramaic prefix de- but in his commentary on the Talmud and in Siddur Rashi he did use Aramaic and not only technical terms; his language in this respect can be described in the same terms as we shall use for the language of rabbinical Hebrew in general. Rashi created new words and patterns; his understanding of Hebrew grammar and his ability as a stylist give him a special place among the writers of Hebrew in Ashkenaz.

ASHKENAZIC AND RABBINIC HEBREW

The Jews of Ashkenaz (western Germany and eastern France) had close ties with Ereẓ Israel, and this relationship is very evident in their piyyutim (Gershom ben Judah Me'or ha-Golah, Jacob Tam, Meir of Rothenburg etc.), which continued the language and grammar of Palestinian liturgical poetry, though on a more modest scale. However, in the responsa, in the books on ritual, in community records, and to some extent, also in books dealing with their trials and tribulations, there is apparent, already from the 11th century, the beginnings of the blended style known as rabbinical Hebrew, found in its most characteristic form in the responsa written in Poland, mainly from the 16th century. It is composed of mishnaic Hebrew, biblical Hebrew, Aramaic (largely from the Babylonian Talmud), a certain amount of Arabic-influenced Hebrew, the influence of Middle High German in Ashkenaz, and once the center of Ashkenazi Jewry had moved to Poland the influence of Yiddish, whose German component was the same kind of German as that spoken by the Jews of Ashkenaz.

The Influence of German on the Hebrew of the Jews of Ashkenaz and the Influence of Yiddish on Hebrew in Poland

The status of Hebrew in Ashkenaz as compared with Middle High German was different from its status in Spain compared with its sister-tongue Arabic. In Germany the two languages in contact were from different families and far apart in form and structure. It is the strangeness of the effect of the influencing language that is most marked, though the very distance can also tone down the influence. The Ibn Tibbons were well aware that it was the closeness of Arabic to Hebrew which secured it such huge influence. From Arabic Hebrew borroweda few words in Hebrew declensions; from German, at the beginning of the period, no words were borrowed at all, and they were quoted as foreign whenever they were needed in explanation: "כְּוִיָּה or in German brennt" (Eleazar b. Nathan, 11th century). Loan words begin to appear in Hebrew in Poland from the 16th century, more in the questions that were posed than in the responsa of the rabbis, and almost all of them dealing with everyday life, hardly any concerned with matters of ritual. They include names of colors and clothes, food, and diseases. Words were also borrowed for which Hebrew equivalents existed: Diamant (יַהֲלֹם), Juwelen (תַּכְשִׁיטִים) etc. It is the language of the Ḥasidim and the Mitnaggedim in the 18th century which is most full of loan words.

Whereas Spanish Jewry was bilingual as far as writing was concerned (Hebrew and Arabic), in Ashkenaz Hebrew served as the sole written language. The literary language of the surrounding culture was mainly Latin, though Middle High German was also beginning to be used in writing for epic poetry, courtly lyrics and sermons. The Middle High German used in sermons to bring people to confession, repentance and fear of sin could well have influenced Ḥasidim of Germany by virtue of the subjects themselves but only as a literary language heard in sermons out of doors, not as a written language. Its influence should therefore be considered as that of a spoken vernacular not as a vehicle for literary expression. The Jew did not regard it as an enlightened, respectable language, worthy of competition with such an excellent tongue as Hebrew. They did not imitate it, they did not translate from it, and they had very little occasion to adopt from it terms and concepts that needed a Hebrew guise. The main effect is felt in passages dealing with everyday life: the account books of the religious congregations, responsa dealing with everyday affairs, accounts of troubles and persecutions.

When Polish Jewry replaced German Jewry as the spiritual center, a change began to take place in the status of Yiddish, transferred to a Slavic environment (though there had been earlier written documents in Yiddish); it became henceforth a normal second written language. Among the responsa written by Moses Isserles, Solomon Luria and others in Poland in the 16th century there are also some written in Yiddish. No study has yet been made on the relation of Hebrew and Middle High German in Germany as compared with that of Hebrew and Yiddish in Poland. Generally speaking, the German element in the Hebrew of Ashkenaz is close to that in the Hebrew of Poland; it is only in the language of the Ḥasidim and the Mitnaggedim in the 18th century that the influence of Yiddish is far more profoundly felt.

In the history of Ashkenazi Hebrew, a special place is reserved for the language of Sefer Ḥasidim, a collection of tales and customs attributed to *Judah he-Ḥasid from Regensburg and written or collected by his disciples in the 12th and 13th centuries. The spirit of modesty and humility typical of those ascetic God-fearing Ḥasidim permeated not only the subject-matter but also the language, which was very close to the spoken variety, abounding in anacolouthon, unstylized, without interlarded quotations or figurative embellishment. The vocabulary of the book is small, sufficient for the needs of the subject-matter, with no concern for the needs of style, and the influence of Middle High German is quite strongly felt. The influence of Middle High German and the German element in Yiddish is best described by reference to the Sefer Ḥasidim and the responsa written in Germany and Poland. The following are some of the most noteworthy features: there is considerable use of prepositions in the German manner, though Hebrew prepositional usages were not rejected completely; they are also used, the proportion depending on the writer, the translator or the context (as already noted, the influence of German is more marked in passages dealing with everyday affairs: -גָּנַב לְ instead of מתגָאה מִן … ,גָּנַב מִן ("proud of"') as in sich rühmen von; יָדַע מִן as in wissen von. The preposition אַחֲרֵי takes on the uses of nach – שָׁלַח אַחֲרָיו ("send for him"), and so on. The preposition עַל is used like auf: קָנָה עַל הַשּׁוּק ("bought in the market") auf den Market; הָיָה עַל הָרְחוֹב ("was in the street") – auf der Strasse; הִמְתִּין עַל פְּלוֹנִי ("wait for someone") – auf jemand; and even more in the language of the Ḥasidim in the 18th century: נָסַע עַל שַׁבָּת ,קֶמַח עַל פֶּסַח etc. Though the possessive construct pattern found in the Bible is commoner than the prepositional structure with šel found in mishnaic Hebrew, there are exceptional cases where a prepositional structure with min is used, in imitation of von: גַּבַּאי מִן הַצְּדָקָה ("a collector of charity"). There are a few loan translations, mainly for the purpose of expressing concepts from daily life: בֵּית הָעֵצָה (Rathaus); בֵּית עֲנִיִּים ("poorhouse"); בֵּית יְתוֹמִים ("orphanage"). Much commoner is the extension of the meaning of the Hebrew word in accordance with the meanings of its Middle High German equivalent: נֶגֶד or כְּנֶגֶד means "approximately," a secondary meaning of gegen, and in later Hebrew -סָבִיב לְ is used in this sense, like um. יָדַע means "be able" as well as "know," since koennen has both meanings; הַפָּסוּק עוֹמֵד כָּתוּב (as in Ger. steht), תּוֹפֵס means "hurry" (chappen), עָזַב means "let" (lassen), אֵיזֶה means "some" as well as "which," both meanings of welche; לְהוֹזִיל means "agree" (from billigen); חוֹלֶה אֵצֶל הַקּׁר "sick because of the cold," a secondary meaning of Middle High German bi (the similarity between אֵצֶל and עַל יְדֵי may have helped). In the responsa, the phrase הָיָה לוֹ expresses duty and obligation, not permission and ability as in mishnaic Hebrew and Aramaic; this is the meaning of Middle High German hân ze. Very common is the use of מַה as a relative pronoun – מַעֲשִׂים מַה שֶּׁחָפֵץ שֶׁיַּעֲשֶׂה "deeds that he wants to perform" (Sefer Ḥasidim) – like Middle High German was. One new sentence structure came from Middle High German: a subjective clause whose predicator was a modal adjective (אָסוּר "forbidden," צָרִיך "necessary," טוֹב "good") could begin with the word אִם, which gave it the force of a conditional clause – יֵשׁ דְּבָרִים שֶׁאָסוּר אִם יְלַמֵּד לַאֲחֵרִים "there are things which it is forbidden that (or "if") he teaches (them) to others." A similar structure with ob is found under such conditions in Middle High German.

German also produced a marked increase in the use of features which already existed in Hebrew though less conspicuously. רַק in the prepositional sense of "but" is found in the Bible אֲשֶׁר לֹא תְדַבֵּר אֵלַי רַק אֱמֶת ("tell me nothing but the truth"; I Kings 22:16) – and this biblical stylistic feature occurs in Spanish Jewish poetry. In Ashkenazic Hebrew רַק is also used as a conjunction in the sense of "but, however," before a verbal clause; it also occurs in the form רַק שֶׁ presumably under the influence of Middle High German nur, which means both 'but' and 'only.' An example is חָס לִי לִרְאוֹת כָּזֶה וְלֹא לִמְחוֹת וְרַק כָּל זֶה עֲדַיִן הוּא שׁוֹרֶשׁ רׁאשׁ וְלַעֲנָה "God forbid I should see such a thing and not protest, but all this is still a source of evil and corruption" (from the 17th century Shenei Luḥot ha-Berit). עַד occurs with the meaning "as long as," found in biblical and mishnaic Hebrew, and assisted by the fact that bis carries this meaning. In the Bible אָז sometimes comes at the beginning of the apodosis in a conditional sentence: לוּלֵי תוֹרָתְךָ שַׁעֲשֻׁעָי אָז אָבַדְתִּי בְעָנְיִי (Ps. 119:92) "if thy law had not been my delight then I should have perished in my troubles." In Ashkenazic Hebrew the high frequency of אָז after a conditional or temporal clause can be attributed to the corresponding use in German of denn, e.g., in the language of Berthold from Regensburg, a preacher at the time of the Sefer Ḥasidim. The normal pattern is thus אִם יִהְיֶה בְּהַגְרָמָתוֹ אָז כָּל הַקָּהָל פְּטוּרִים "if he is the cause, then the whole congregation is exempt."

The large number of expressions in German with machen is matched in Ashkenazic Hebrew by the number of expressions with עָשָׂה, though a few such phrases can be found in the Midrashim: עָשָׂה רְצִיחָה עָשָׂה חֶרְפָּה ,עָשָׂה נְזִיפָה ,עָשָׂה עַוְלָה ,עָשָׂה עֲבֵרָה, etc. The influence of German also explains the number of expressions like הָיָה לוֹ עָוֶל ,הָיָה לוֹ בּוּשָׁה ,הָיָה לוֹ צַעַר. Here it is the frequency which is affected, since they are not a complete innovation in Hebrew. The use of sich as a reflexive object leads to the use of inflected עֶצֶם in Hebrew translation – equivalents; to the few examples found in mishnaic Hebrew are added מְסַגֵּל עַצְמוֹ ("adjust oneself"), מַחֲזִיק עַצְמוֹ ("hold oneself," i.e., avoid, as in Ger.) תּוֹפֵס עַצמוֹ ("consider oneself"). It was also used with verbs in binyan hitpaʿẹl, which in such cases lost reflexive meaning: מִתְקַשֵׁט עַצְמוֹ ("adorn oneself"), מִתְלַבֶּשֶׁת עַצְמָהּ כְּכוֹמֶרֶת ("dresses herself as a priestess" – Sefer Ḥasidim), etc.

German could also lead to a diminution in frequency of occurrence: the relative infrequency of אֶת as compared with biblical and mishnaic Hebrew can be explained by the absence of any corresponding particle in German. Similarly the reduction in the use of the definite article corresponds with its reduced use in Middle High German; נָטְלוּ כּׁל and not… הַכּׁל is modeled on the use of alles.

No new tense forms were created in Ashkenazic Hebrew, but to some extent the systemic relationships of existing forms were reorganized (see Rabin in bibl.). In Sefer Ḥasidim the use of the participle for both present and future indicative is well marked: לֶעָתִיד לבוֹא מִתְפַּלְּלִים. The imperfect יִקְטׁל serves as present and future subjunctive, expressing doubt, possibility etc.: הַזֶּרַע שֶׁל גֵּר יִהְיֶה צַדִּיקִים "the seed of a proselyte may produce righteous men." And this systemic relationship between קוֹטֵל and יִקְטׁל corresponds with the opposition between present-future indicative and present-future subjunctive in Middle High German, where the expression of futurity by means of modal auxilliaries – will, soll – was still rare. Moreover, since there is a firm foundation for this division of function between קוֹטֵל and יִקְטׁל in mishnaic Hebrew and Aramaic, the novelty is felt more strongly in the use of the pattern הָיָה פּוֹעֵל not only as a past continuous (a usage inherited from mishnaic Hebrew) but also as a functional, though not formal, equivalent of the past subjunctive in Middle High German: וְהָרַב לֹא הָיָה מְלַמְּדוֹ בְּחִנָּם "the rabbi would not teach him for nothing" (Sefer Ḥasidim, section 585). This use is also not a complete innovation, since it resembles the combination הָיָה and present participle in unfulfilled conditions in Aramaic and in mishnaic Hebrew (אִלּוּ הָיִיתִי יוֹדֵעַ "if I had known," הֲוָא אמינא); what is new is the systemic relationship between קָטַל and הָיָה קוֹטֵל, corresponding to the past indicative and past subjunctive in Middle High German. There are examples in Ashkenazic Hebrew of the sequence of tenses found in German: רָאָה בַחֲלוֹם שֶׁהָיָה רוֹכֵב עַל סוּס אָדׁם "he saw in a dream that he was riding on a red horse" (in Sefer Ḥasidim); the tense in the subordinate clause is marked as past, like the main clause, and not present, which is normal in Hebrew.

The Role of Mishnaic Hebrew in Ashkenazic Hebrew

The language of Jews who studied Talmud naturally made great use of mishnaic Hebrew. It should be noted that Sefer Ḥasidim actually uses forms that are not particularly common, e.g., binyan nupʿal – נֻצַּל ("was saved"), נֻטַּל ("was taken" – and the pattern הַקְטִילָה as verbal noun for binyan hipʿil – הַפְסִידָה ,הַרְעִיבָה. The most productive derivational pattern by far was קַטְלָן from mishnaic Hebrew; the innovations include עַצְרָן ,פַּשְׁרָן ,שַׁדְּכָן ("miser"), שְׁתַדְּלָן ,מַטְבְּעָן ("coiner"), עַוְלָן ,שַׂמְחָן, etc.

The Role of Biblical Hebrew

Ashkenazic Hebrew took from all varieties known to its writers; the Hebrew of the Bible was absorbed from the stratum of biblical Hebrew in the liturgy, from the weekly readings of the law and from the hafṭara. Hence the use of יַעַן ,טֶרֶם אָז ,יִתָּכֵן ,לְמַעַן, and so on. And unlike Arabic-influenced prose, Ashkenazic Hebrew also made use of lengthened tense forms: אֶכְתְּבָה ,יִשְׁמְרוּן and the future with waw conversive. This tense is seldom found in Sefer Ḥasidim and relatively infrequently in the prose of Rashi, but it began to occur with increasing frequency until it became a distinguishing mark of rabbinic Hebrew. The increasing use of rhetorical figures from the Bible may have helped to establish it in the language; fragments of verses which contained a waw conversive were directly quoted as part of the rhetorical figure, and thus made their way into rabbinic Hebrew. Biblical figures of speech are absent from the humble style of Sefer Ḥasidim, and used with taste and moderation by Rashi; in the language of Jacob Tam they are widespread (their flavor of piyyut derives from his being a liturgical poet), and they are quite common in the writings of Rabbi Meir of Rothenburg (end of the 14th century), for example. They are very common in the salutations, at the beginning of many of the responsa from Germany and Poland, though not only there. Some writers use them more, some less, depending on their individual taste and ability. Though most rhetorical figures derive from verses in the Bible, there are also some from mishnaic Hebrew, and even from Aramaic, all for the rhetorical adornment of the opening section of the letter, with no conception of the principle of purity of biblical language as a rhetorical virtue. The writers of the responsa begin שָׁמַעְתִּי וַתִּרְגַּז בִּטְנִי ("I heard and my stomach quaked"; Meir of Rothenburg); or צָלַלְתִּי בְּמַיִם אַדִּירִים וְהֶעֱלֵיתִי חֶרֶס (see Ex. 15:6, "I plunged into deep water and brought up nothing"; Moses Isserles); or וָאֶתֵּן אֶת לִבִּי לִדְרשׁ וְלָתוּר סִבּוֹתָיו בַּסְּפָרִים (see Eccles. 1:13, "I gave my heart to seek and search out its reasons in books" (a frequent figure)). They begin with a rhymed eulogy: (see Ps. 18:30, אִמְרָתְךָ צְרוּפָה וּמַרְחַשְׁתְּךָ עֲמֻקָּה וְצָפָה "thy wordis perfect and thy feelings are deep…"; Meir of Rothenburg)). Sometimes the point of the quotation distorts the original meaning of the verse quoted: Simeon accuses Reuben in the words בִּפְלַגּוֹת רְאוּבֵן גְּדוֹלִים חִקְרֵי לֵב (see Judg. 5:16, "in the divisions of Reuben there were great heart-searchings" (a question addressed to Jacob Tam). Sometimes the form is changed: the Rabbi replies עִרְעוּרִים וְקוֹל (not יַעֲקׁב) הַקּוֹל קוֹל עָקׁב – "the voice is the voice of the deceit." Sometimes the spelling of the word is changed: הָאֶרֶז אֲשֶׁר בַּלְּבָנוֹן בָּאֵר ("the cedar of Lebanon is proof") the reference is to Isaiah 40:16 where the form is actually בָּעֵר ("burns"; a quip from the responsa of Šaʿar Efrayim, 17th century).

The Role of Aramaic

In Sefer Ḥasidim there are slight touches of Aramaic – כִּדְבָעֵי דִּכְתִיב, – and they are fairly restricted in Rashi's commentary on the Bible. However, in Rashi's commentary on the Talmud, in the responsa of Jacob Tam, or in Sefer ha-Rokeaḥ of Eleazar of Worms, the amount of Aramaic acquires a status comparable to that of mishnaic Hebrew, since both constitute the language of the Talmud and the writers may not always have realized when they had moved from Hebrew to Aramaic.

The extensive use of Aramaic is not confined to the Aramaic halakhah under discussion, nor even to the technical terms alone (קָיְימָא לָן ,אַלְמָא ,טַעֲמַיְהוּ ,צְרִיכָא ,פְלִיגֵי, etc.). -דְּ is used instead of שֶׁ in an otherwise completely Hebrew context, to mean "of" – הַפֶּרֶק דְּהַמַּסֶּכֶת (the chapter of the tractate); הָרַב דִּשְׁפֵירָא ;פִּנְקָס דְּוַעַד אַרְבַּע אֲרָצוֹת – and also to mean "that" – -נִרְאֶה לִי דְּ ("it seems to me that…"); אָמַרְתִּי דְּאָסוּר ("I said that it was forbidden"). Whereas as Rashi commenting on the Bible writes -אַף עַל פִּי שֶׁ- ;כֵּיוָן שֶׁ- ;מִשּׁוּם שֶׁ when he comments on the Talmud he writes -אַף עַל גַּב דְּ- ;משּׁוּם דְּ- ;כֵּיוָן דְּ The verb forms מַיְירי and מַיְיתי are much used in Hebrew contexts, and sometimes complete clauses in Aramaic are interpolated. Generally the Aramaic phrases are quoted freely in new Hebrew contexts, though in the form and with the inflections found in the Talmud. The Hebrew most thoroughly mixed with Aramaic was written in Poland during the 17th and 18th centuries.

The Link With Arab-Influenced Hebrew

The Jews of Ashkenaz and the Jews of Spain maintained cultural ties, whether by means of responsa (e.g., when Asher ben Jeḥiel moved from Germany to Spain) or by reading each other's books. The kabbalists of Spain were interested in the Ḥasidism of Ashkenaz, in Germany they read the writings of Maimonides, Saadiah Gaon, and the musar book, Ḥobot ha-Leḇaḇot. In Poland several of the scientific works, all written in "Tibbonian" Hebrew, were well known.

Since Ashkenazic Hebrew reflects all the varieties of literature known to its writers, and since in principal there was no form of the language whose use was banned on stylistic and grammatical grounds, it is not surprising that several features of Arab-influenced Hebrew also occurred. The use of nouns ending with the suffixes ית-. and וּת- in the masculine is characteristic of rabbinic Hebrew (פְּרִישׁוּת רַבָּנִי ,תַּכְלִית אַחֲרוֹן), and similarly plural concord of the form סוֹדוֹת גְּדוֹלוֹת (instead of גְּדוֹלִים). It is not clear whether this usage derives from Spanish Hebrew or whether it is an aspect of the general weakening of strict grammatical rules (it is possible that this is due to the fact that words ending in -ut and -it in Yiddish entered the neuter gender). Sometimes Arabic and German/Yiddish tended to produce the same result, e.g., the demonstrative without the article before the noun, reduced use of אֶת, the preposition מִן in סִפֵּר מִן ,דִּבֵּר מִן and the comparative form יוֹתֵר נָאֶה ,יוֹתֵר טוֹב, where the frequency and the order of words can be attributed to German (mehr…) and French (plus…), though Arab-influenced Hebrew also uses this structure as a translation of Arabic.

A deliberate, quite marked use of the Hebrew which took shape in Spain and Provence is found in Poland, mainly from the 16th and 17th centuries. Moses Isserles was a man of philosophic bent and apart from his responsa on matters of halakhah he also used the language of the Guide of the Perlexed: אַקְלִים ,אֱלֹהִי ,עִיּוּנִי ("region"), מַסְכִּים ("match, fit"), בָּרַח מִלַּעֲסֹק ,מֻשְׂכָּל ("refrain from doing"…), עַל צַד הֵשּׂכֶל and so on.

Forty years after the expulsion from Spain 16th-century Safed became a center for Jewish learning and Kabbalah whose greatest scholars generally wrote a Hebrew close to the Arabic-influenced variety of Spain and Provence, especially in Kabbalistic works where a style of writing had already been established in Spain (see, for example, the language of Moses Cordovero). In Safed, too, the use of lengthened tense forms and the waw conversive were introduced into the language of prose. Isaiah Horowitz, author of Shenei Luḥot ha-Berit, was educated in Poland and wrote an Ashkenazic Hebrew, but his style contained many of the distinguishing marks of Arabic-influenced Hebrew: the definite article before an infinitive e.g., רֹשֶׁם ,הַהִשְׁתַּמֵּשׁ for "influence," מַסְכִּים ,מֻשְׂכָּל meaning "match," אֶמֶת in the masculine, מַהוּת ,עַצְמוּת and so on; after all, he was a kabbalist, immigrated to Ereẓ Israel, and wrote his book there.

The extreme case of the encounter between Spanish Hebrew and Ashkenazic Hebrew is the language of the 18th century Ḥasidim. Besides being stamped with the imprint of Yiddish to a greater extent than any preceding variety of Hebrew, it also continues the traditional prose style of Ashkenazic Hebrew (as exemplified in the responsa, in the musar books and especially in the well-loved Sefer Ḥasidim). But whereas stories of the ẓaddikim and passages dealing with everyday life are written mainly in Ashkenazic, rabbinic Hebrew, the philosophic literature of Ḥasidim is strongly marked by Spanish Hebrew; the ḥasidic writers continued the kabbalistic tradition of Isaac Luria, and took over the terms and expressions from kabbalistic literature and Spanish books of ethics such as Ḥobot ha-Leḇaḇot; רוּחְנִיּוּת ("spirituality"), אַחְדוּת ("unity"), גַּשְׁמִיּוּת ("corporeality"), בְּתַכְלִית, דְּבֵקוּת, סְגֻלָּה ,בְּחִינָה ,etc. (see M.Z. Kaddari in bibl.).

Influence of Hebrew on Yiddish

It was through rabbinic Hebrew, with its blend of all varieties, that Hebrew words found their way into Yiddish. From biblical Hebrew – עַוְלָה אָז יִתָּכֵן; from mishnaic – אֶפְשָׁר ,אֲפִלּוּ; from Aramaic – פְּשִׁיטָא ,סְתָמָא, and even from Arab-influenced Hebrew, mainly via ḥasidic literature: הֶעְדֵּר שְלילָה ,מְגֻשָׁם ,גַּשְׁמִיּוּת, etc. Many of the high frequency words most characteristic of Ashkenazic Hebrew, words occurring already in Sefer Ḥasidim, came into Yiddish: מֵאוּס ("obnoxious"), עוֹלָם (with the meaning "people," as in Aramaic קָהָל ,(כֻּלֵּי עָלְמָא ("an urban Jewish community"), מַעֲשֶׂה ,מַמָּשׁ (in the language of the Sefer Ḥasidim it already has the meaning "story" as well as "deed"), אָז ,אוֹדוֹת ,תָּדִיר. The Hebrew derivational pattern most characteristically Yiddish – קַטְלָן – (in words like בַּטְלָן ,בַּלְעָן ,יַקְרָןיַשְׁרָן, etc.) is also the pattern most vital to Ashkenazic Hebrew.

Haskalah and Medieval Hebrew

The stylistic uniqueness of rabbinic Hebrew lies in its blend of different varieties of the language: only a few new words were coined, to meet the needs of writing about everyday life. By contrast, Spanish Hebrew was a professional tool, a necessary instrument for all kinds of scientific, philosophic and scholarly writing. The writers of the Haskalah turned their backs on rabbinic Hebrew for its careless grammar and because it represented the Judaism of the Talmud. For poetry and to a considerable extent for stories they adopted biblical Hebrew; however, for serious prose works some of the maskilim chose the Arabic-influenced language of Spain, especially for technical terms and expressions (see for example the extensive use of Tibbonian Hebrew made by Naḥman Krochmal in Moreh Nebukhei ha-Zeman).

The Hebrew language was a major concern of Haskalah writers. They were keenly aware of normative problems in writing and the need for linguistic research in Hebrew. Writers and grammarians like Naphtali Herz *Wessely, Judah Leib *Ben Ze'ev, and those who collected ancient texts made a decisive contribution toward the molding of the language and its modernization. The Haskalah may be seen as a preparatory period for the revival of Hebrew (see Modern *Hebrew Literature).

[Esther Goldenberg]

MODERN PERIOD

The growth of Hebrew as a modern language, spoken by masses and gradually used in all areas of life and thought, may be divided into three stages corresponding to periods in the history of modern Palestine: (1) 1881–1918 initiated by Eliezer *Ben-Yehuda's arrival in the country. He and his followers developed and propagated Hebrew in everyday life. (2) 1918–1948; under British rule when Hebrew was first considered a language of Palestine, and later (1922) one of the three official languages. During this time the Hebrew-speaking population increased rapidly, established many cultural institutions, including its own educational system up to university level, in all of which Hebrew, with few initial exceptions, became the only language used. (3) 1948– marked by the foundation of the State of Israel. Hebrew became the predominant language of the state, and was used in all branches of its activities: government departments, the army, etc., were integrated into the life of the Hebrew-speaking population. Gradually Hebrew was also spoken by non-Jewish citizens. Each of these three periods, characterized by the cultural background and the linguistic past of the immigrants who adopted Hebrew as their new language, has influenced its revolution.

[Eli Eytan]

At the time of the revival of Hebrew for everyday speech, the languages most current in the old yishuv (Jewish population) were *Yiddish, *Ladino, and *Arabic, while French and German formed the main channels to European culture. The immigrants of the first period, mostly from Eastern Europe, spoke Yiddish; many of them also spoke Russian or Polish and at least understood German. These languages influenced Hebrew but their effect, noticeable in new aspects of Hebrew, gradually decreased, and the impact of English grew. Since the end of World War I English had a marked influence on Hebrew because of the influx of British and other English-speaking government and army personnel and their closer contacts with the yishuv. The fact that the establishment of the State of Israel did not diminish this influence is due to a wide knowledge of English among the Israeli population through higher education and close acquaintance with English and American culture to which immigration from English-speaking countries contributed substantially.

Period of Revival (1881–1918)

Hebrew was spoken in Palestine even before the revival movement, but only as a lingua franca among Jews who had no other common language. This phenomenon also existed among Jews in many other countries in earlier periods. The revival, in contrast to early periods, however, saw the establishment of Hebrew as the sole or at least the principal language, i.e., a transformation from a language used only occasionally for special purposes by speakers of other languages to a language used by a community for all their communication needs – speaking, reading, and writing.

The revival took place in Palestine. When the British conquered the country, Hebrew was already one of the languages of Palestine. In General Allenby's published proclamation about martial law in Jerusalem, Hebrew was published on top, while Arabic was the second, before Russian and Greek, all considered languages used by the local population. On the other side of the sheet, the proclamation was published in (1) English, (2) French and (3) Italian, languages of the allies. Only toward the end of this period, Hebrew also began to be studied in the Diaspora to a limited extent.

Eliezer Ben-Yehuda's pioneer work for the revival of Hebrew would have failed had there not been at that time three conditions which proved essential to the process of revival: (1) There was no national language in Palestine. The inhabitants did not belong to a "nation" (in the Western sense), but were divided into religious-ethnic communities ("millets") that used a number of languages. Literary Arabic was the language of prayer, worship, and study for all Muslims including government workers and members of the Turkish army and, to some degree, for several Christian denominations. But millions of Muslims outside Palestine also used literary Arabic in a similar fashion, thus preventing it from being an exclusive national language. Turkish was used for political, governmental, and military matters all over the Ottoman Empire. The most common spoken language was local colloquial Arabic, which was used only as a spoken vernacular and thus deemed unworthy to be a national language. Other languages, such as French, Russian, Italian, Greek, and Armenian were used by certain millets, or as cultural languages. None of them, however, could be taken as a national language. (2) European nationalist thought, together with a yearning for a Hebrew renaissance, reached Palestine in the middle of the 19th century. Already in the 1860s young people in the Jewish communities of Palestine attempted to change the static way of life there. Newspapers, printing houses, and various workshops were founded, and settlements were established "outside the walls" (i.e., of the old yishuv). The lack of a national language in Palestine created the need for a common language for the developing society, and it was natural that Hebrew be considered worthy of this role; all the more so that Hebrew (in the Sephardi pronunciation) even before this had been the common language of different Jewish communities. (3) The fact that the original language of the country had been Hebrew provided a solid ideological basis for the revival of the language, and gave it an advantage which no other language had. Publication of such ancient Hebrew inscriptions or engravings as the Siloam inscription (1880) and the Mesha inscription (1868) made a deep impression upon the people of that generation and emphasized the connection between Hebrew and Palestine. The revival of the language symbolized the "Golden Age" of ancient Israel which was about to be renewed.

The major difficulty encountered in making Hebrew the sole (or principal) language of the country was in the area of vocabulary. There were few difficulties, if any, in the field of grammar. In Hebrew phonology the need for marking such new sounds as č, ž, ǧ (to accommodate foreign words and non-Hebraic personal names) was met without difficulty by adapting the letters ג׳ ,ז׳ ,צ׳. (These sounds had previously been marked by combinations of letters such as ז׳ .דזש ,זש ,טש had been indicated for some time also by the letter י׳, undoubtedly through the influence of the French pronunciation of the letter j.) The problems of orthography were solved at once: "defective" orthography (כתיב חסר) was introduced. There were, certainly, difficulties in this area, and it is relevant to mention Ben-Yehuda's short-lived experiment in the use of "capital" letters for personal names (as in English and French). Morphology was not expanded, but newly invented words were usually styled according to existing morphological patterns. It is often possible to distinguish tendencies to use a certain pattern or a specific suffix, such as Ithamar *Ben-Avi's predilection for the suffix of relation (יהודה מיידית ,עמדה זכותית). Although Hebrew syntax changed considerably during the days of the revival of the language, these changes were generally brought about unintentionally and without premeditation. (However, an apparent example of an intentional syntactical change is to begin sentences with a verb, like in Arabic, as was done for a time in newspapers.) In contrast, the need for new words was recognized from the start. Ben-Yehuda illustrates this in the following statement: "Have any of the readers (of Smolenskin) ever felt that in all of the circumstances of the different events that this very capable author brought into his stories, he never mentioned for example, the simple, common act, of tickling? This act which we meet often in every story in a living language we will never meet in the stories of Smolenskin, simply because he did not have a word for it. In spite of this his stories are well written. But whoever wishes to write something of wisdom and science, and especially someone like myself, who speaks Hebrew at home with the children, about everything in life, feels every moment a lack of words without which living speech cannot take place" (the Large Introduction," 12–13). Most of the efforts of those who revived the language were dedicated to answering this need.

The End of the Revival Period

The period of revival was characterized by reviving existing words, creating new ones, and enriching the language with words from Semitic sources (in the main) cast in the Hebrew mold. However, a large number of these words (several thousand) were rejected and have fallen into disuse. The pressing need to remedy the critical lack of words often led to hasty innovations. Those educated in literary Hebrew, especially the last generation of maskilim in Eastern Europe, did not readily accept this "manufacturing of words" in Palestine. They tended to be more careful in making innovations, preferring to adopt foreign words, especially "international" terms, the majority of which were of a Latin or Greek origin. This school of thought began to make its influence felt in Palestine from 1905 onward, with the Second Aliyah. The coming of the Third Aliyah from Eastern Europe, immediately after World War I, strengthened this view until in the late 1920s the influence of the "language of the revival" could hardly be recognized since many of its words were forgotten. The late books and journalism of Ithamar *Ben-Avi were a kind of "swan song" of the revival period, but even his language greatly reflected the abovementioned changes. The end of the Ottoman Empire and the recognition of Hebrew as an official language in Palestine is therefore only one reason for fixing the end of World War I as the close of the "revival period." The other reason is that at this time the influence of those who demanded great caution in the formation of new words grew, and they were tolerant to foreign words as long as proper Hebrew terms had not been created with careful consideration.

See also Y.M. *Pines, Z. *Jawitz, the *Academy of the Hebrew Language.

[Uzzi Ornan]

Linguistic Problems of Modern Hebrew

PRONUNCIATION

Reviewing the first 22 years of the Va'ad ha-Lashon (Zihkronot Va'ad ha-Lashon I, 2nd ed. p. 4), Ben-Yehuda recalls the days when all the various pronunciations of Hebrew were heard in Jerusalem "from the Lithuanian to the Sephardi, from the Volhynian to the Yemenite and the Persian." The necessity to establish a standard pronunciation was under discussion for some time. At a meeting of teachers in 1885, for example, it was decided to teach Ashkenazi Hebrew for the first two years in Ashkenazi schools and then switch to Sephardi pronunciation, while in Sephardi schools the opposite should be done – in "order that they know both." By 1912, however, Ben-Yehuda continues, "by the nature of things the Oriental pronunciation, the one living among the Sephardim, had become dominant, and from Jerusalem it spread to all speakers of Hebrew in the country." This statement was a rough summary of the position which had developed in a relatively short time, but which, in fact, was – and is to this day – only a limited fulfillment of the original wish to adopt the Sephardi pronunciation. The phonetic inventory of the Ashkenazim, both of the old yishuv and new immigrants whose number rapidly grew, could not easily be replaced by the whole range of sounds found in genuine Sephardi speech. To give one example, any Ashkenazi could easily learn to pronounce both vowels of the verb עָצַר ("he restrained," etc.) as a and to stress its last syllable, instead of pronouncing the first vowel as o (as according to his Ashkenazi dialect), and stressing the first syllable. It proved, however, impossible for the vast majority of Ashkenazi Jews – except by sustained conscious effort to – pronounce the Sephardi (like Arabic) consonants ע [ʿ] and צ []. עָצַר and אָצַר ("he stored up"), as well as many other pairs of linguistic forms, thus became homophonous, creating new problems in teaching orthography and grammar. This process of different phonemes coinciding in actualization even led to certain restrictions in the use of the existing vocabulary and in the possibilities of its enlargement. אָצַר, e.g., is hardly ever used in everyday speech, and a possible new noun מַאֲצָר ‡ would be rejected owing to its homophonous rival מַעֲצָר ("retention, arrest"). The difficulties, stemming from the homophony of originally distinctive features, constituted, and still do, one of the main arguments to continue trying to propagate a purer Oriental pronunciation. On the other hand, the common "Sephardi" pronunciation had meanwhile acquired a certain value of social superiority, since most of the leaders of the new yishuv came from Ashkenazi circles, and many Oriental Jews, whose original speech did contain the sounds in question, abandoned that part of their native phonetic inventory in order to imitate the speech of their social superiors.

At the first convention of the Hebrew Teachers' Association in 1903, the pronunciation issue was discussed, but no decision was taken, mainly because Eliezer Ben-Yehuda and David Yellin, members of the Va'ad ha-Lashon, held different views on the desirability and feasibility of one or the other feature of genuine Oriental pronunciation being adopted as standard. The debate went on until, in 1923, the Va'ad ha-Lashon decided to demand the following reform in the pronunciation of the Hebrew letters in question: ב without dageš = English v; ו = Arabic wāw and English w; ח = Arabic ḥā as distinct from כ without dageš = Arabic ḫā and German ch (as in "Bach"); ט = Arabic ṭāʿ emphatic t; ע = Arabic ʿayn; צ = German z (tz); ק = Arabic qāf (emphatic velar); ת without dageš = Arabic and English unvoiced th (as in "thin").

It is noteworthy that this ruling does not follow the Sephardi tradition in all details. Here, both alternants of bet, one written with dageš (בּ) and the other without (ב), were pronounced by some Sephardi communities as b, and the settlers of Galilee, following their Sephardi teachers, had already adopted this pronunciation. The Sephardim also pronounced צ as emphatic unvoiced ṣ, like Arabic ṣād, while the pronunciation tz provided for in the ruling is Ashkenazi. Furthermore, nothing was said about the vowel segol (־ֶ), which in Sephardi speech is not distinguished from ṣere ־ֵ), i.e., closed , while Ashkenazim and some Israeli speakers to this day distinguish segol from ṣere. The Va'ad ha-Lashon in this decision also omitted mentioning the difference in pronouncing qameṣ preceding ḥaṭaf-qameṣ, as in נָעֳמִי, which the Sephardim pronounce Naʿomi and the Ashkenazim (and most Israelis) Noʿomi.

The authority of the Va'ad ha-Lashon was not sufficient to enforce this reform in the face of already-established speech habits. Consequently, current Hebrew pronunciation differs from that of the Ashkenazim in the following details only: (1) qameṣ-gadol (-) = a; (2) ḥolem (וֹ) = o; (3) taw without dageš (ת) = t; (4) ultimate stress of most words, while penultimate stress is confined to some classes of words, as in classical Hebrew.

In the matter of stress, however, the Ashkenazi way has led to some more deviations from the Sephardi and classical Hebrew system. Many proper names of persons and places have penultimate stress in everyday speech: ('Raḥel; 'Moshe, 'Shlomo, 'Ḥ fa, etc.), and the retracted accent has become, particularly in childrens' speech, a mark for names of things charged with some affective value: 'glida ("ice cream"), buba ("doll"), etc. Penultimate and antepenultimate stresses are also characteristics of foreign borrowings: komu'nisti ("communist"), rele'vanti ("relevant") etc., notwithstanding their Hebrew suffix -i; integ'raẓya ("integration"), uni'versita or univer'sita; ("university"), etc.

Efforts to propagate a diction based on classical grammar and Sephardi pronunciation were especially made among broadcasters. The question, however, as to what this desirable correct diction entails and what can be attained was debated up to the late 1960s. Up to that time, the *Academy of the Hebrew Language had recommended to the Israel Broadcasting Service to observe the Oriental pronunciation of ḥet and ʿayin, the gemination of consonants with dageš-ḥazaq, as in hassefer (הַסֵּפֶר); dibber (דִּבֵּר); and the šewa-naʿ, as in devarim (דְּבָרִים), katevu (כָּתְבוּ), dibberu (דִּבְּרוּ). This recommendation was followed to some extent. It is doubtful whether the Oriental diction, though preserved by some Jews of the Oriental communities and applied to Hebrew by Arab citizens, can still contribute to a reform in the speech of wider circles. On opposite trends see Bentolila in Bibliography.

SPELLING

In 1929 when the Va'ad ha-Lashon first published its quarterly Leshonenu, the editors stated in their programmatic introduction: "The problem of spelling has not yet been solved…. Some advocate grammatical spelling, others insist on 'full' spelling. This is why the editors have decided to use, for the present, the accepted grammatical spelling and add complete punctuation wherever the reading is doubtful… Thus, we have attained uniformity of spelling without deciding upon the problem itself." What is meant by "full spelling" (כְּתִיב מָלֵא) here is the method of employing, instead of vowel punctuation, vowel letters to supplement the letters admitted in the "accepted grammatical spelling" which, in turn, is a standardized biblical orthography. This system had been proposed by David Yellin and adopted in the summer of 1905 at the teachers' convention at Gederah. Although it has been taught in schools ever since, the debate on the problem never ceased, and actual usage outside, and partly inside, schools went its own way.

According to this system, every word is spelled in one way only, whether vocalization for vowels etc. is added or left out, e.g., בֹּקֶר ("morning"), בָּקָר ("cattle"), בִּקֵּר ("he visited"), and other words having the same three consonants all have the same written form in unvocalized spelling. Their intended reading is revealed by the context only, unless one or more significant vowel points are added to hint at it, e.g., בֹקר ,בִקר ,בָקר. Yellin's system of unvocalized spelling was based on the orthography that, as far as is known, had first been used by the writers of the Haskalah who tried to follow the Bible in all respects. But since the spelling of the words of the masoretic Bible is not uniform, i.e., the same or analogous forms are sometimes written plene and sometimes defectively (without vowel letters), orthography complying with biblical grammar had to be standardized. While the spelling with few vowel letters in fact causes the reader, who knows Hebrew, less difficulty than the inexperienced may expect, it has been under constant attack for other reasons: (1) It is taught in schools, but most writers and printers continue to insert the available vowel letters in the consonantal skeleton of the word in the same way in which Hebrew has been written for many centuries starting even before the rise of vocalization and continuing side by side with vocalized writing down to the present time. (2) It made the language hard to learn for new immigrants, etc., and occasionally caused even fluent readers to stumble. (3) It demanded of everyone a considerable knowledge of grammar or a rare accuracy in diction that would distinguish between long and short vowels, between geminated and ungeminated consonants, etc.

The advocates of "grammatical" unvocalized spelling, mainly teachers and grammarians, also have some weighty arguments to adduce: (1) Their system nowhere contradicts pointed "grammatical" spelling as preponderantly found in the Bible, prayer books, poetry, etc., and taught in schools in conjunction with grammar from which, in their view, it cannot be divorced. (2) They insist upon the ease with which a learner can pass from vocalized spelling to reading and writing texts with the same allowed vowel letters, but without vocalization. (3) They maintain that supplementary vowel letters obstruct the recognition of word roots and thus hamper learners of the language. (4) They emphasize the educational and cultural disadvantage of the simultaneous currency of two contradictory systems. (5) They stress the absence of generally accepted rules for, and the prevailing confusion in the use of, supplementary vowel letters.

This last point is aimed at the fact that many writers add, whenever they see fit, י for i or e; ו for u or o; and – particularly in foreign words – א for a; and use וו for consonantal ו and יי for consonantal י or the diphthong ay. Thus, סִבָּה may be found spelled סבה or שרפה – שְׂרֵפָה; סיבה or קדש – קֹדֶשׁ ;שריפה or שלחן – שֻׁלְחָן ;קודש or אקדמי – אֲקָדֵמִי; שולחן or אקאדמי or עיור ,עיור ,עור – עִוֵּר ;אקאדימי or בניין ,בנין – בִּנְיָן ;עִיוור or דיי ,די – דַי ;ביניין or דאי.

When the Va'ad ha-Lashon published, in 1948, its "Rules for Unvocalized Spelling" (Lěšonénu 16, 82ff.), this was the outcome of over 30 years of deliberations in general meetings of the Va'ad, in committees, and in subcommittees, where frequently also teachers and scholars from outside took part. The various proposals submitted and discussed included suggestions to equip the Hebrew alphabet for the representation of the vowel a and by creating new letters. The use of Latin script for Hebrew was also advocated, as had been done earlier by Ithamar Ben-Avi and Ze'ev V. Jabotinsky. The principles underlying the rules are set forth in the introduction to the draft rules published in 1943 (Lěšonénu 11, 232ff.) and will be summarized here:

The rules must be founded upon the literary sources and the grammar of Hebrew, adapted to modern pedagogical and practical needs, and be acceptable to the public. Therefore, extreme innovations such as the use of א or ע or new letters as vowel signs are to be avoided. The aim is to regularize the spelling actually current and direct it in line with the general tendency of linguistic and cultural developments. For many generations two spelling systems, the vocalized and unvocalized, have existed side by side, and each has its domain of function. But while punctuation by now has fairly well-established rules, in unpointed spelling two contradictory systems compete, one with and the other without supplementary vowel letters; both of them sometimes intermingle in the same text. The evolution of orthography from its beginning to our days tends toward supplemented spelling; unvocalized orthography must therefore be based on it. This is by no means incompatible with grammar and correct pronunciation, for nowadays Hebrew, like any other living language, is naturally learnt by hearing, not from writing. The aim is to facilitate reading, and that is why, whenever supplemented spelling is liable to mislead, it must be dispensed with. Complete consistency is not sought, but this does not mean giving up the formulation of systematic and scientifically founded rules, it rather explains the exceptions recommended by the committee.

The rules themselves submitted for discussion and decision were substantially the same that were later adopted by the Va'ad ha-Lashon in 1948 and again confirmed, with few amendments, by the Academy of the Hebrew Language in 1969. The following words, each spelled without vocalization in accordance with the rules, followed by unvocalized grammatical spelling, and then again fully vocalized will illustrate the principal rules:

השוּלחנוֹת כּוּלם = השלחנות כלם = הַשֻּׁלְחָנוֹת כֻּלָּם; חוּלצה אדוּמה; חלצה אדמה = חֻלְצָה אֲדֻמָּה; בּוֹקר = בקר = בֹּקֶר; מוֹח = מח = מֹחַ; תשמוֹרנה = תשמרנה = תִּשְׁמָרְנָה; עיקר = עקר = עִקָּר; זימן זמן = זִמֵּן; ניתן = נתן = נִתַּן; עלייה = עליה = עֲלִיָּה; זיכּרוֹן = זכרון = זִכָּרוֹן; עירבוֹן = ערבון = עֵרָבוֹן; פּירשׁ = פרש = פֵּרַשׁ; קירוּב = קרוב = קֵרוּב; ייראה = יראה = יֵרָאֶה; ריאתהּ = ראתה = רֵאָתָהּ; עניין = ענין = עִנְיָן; הצייר צייר = הציר ציר = הַצַּיָּר צִיֵּר; בּניי = בני = בָּנַי; חוֹדשיים = חדשים = חָדְשַׁיִם.

Exceptions to the main rules, i.e., classes of words and letter combinations where no addition of vowel letters is allowed, are shown in these examples: תאמר = תֹּאמַר ;קנה = קָנֹה (in these verb forms with quiescent א and ה, there is no וֹ for o); טהרה = טָהֳרָה ;אנייה = אֳנִיָּה ;חכמה = חָכְמָה ;אמנם = אָמְנָם (qameṣ-qaṭan and ḥaṭaf-qameṣ are normally not rendered by נטיוֹת = נְטיּוֹת ;דיוּן = דִּיּוּן ;(וֹ (i left unmarked when preceding -יוּ- or -יוֹ-).

It will have been noticed that this supplemented unpointed orthography still uses the following diacritical points: dots in וּ and וֹ to distinguish them from each other and from consonantal ו; in פּ ,כּ ,בּ to distinguish them from פ ,כ ,ב; and in הּ to mark this letter as the final consonantal h. While recommending this method, the Va'ad ha-Lashon had made it optional considering that the necessary printing types may not be available. The result could have been foreseen: almost nobody used the dotted letters, but wrote ו for both o and u and used וו to mark the consonant w (v), taking advantage of the alternative allowed by the Va'ad. Thus, in fact, בֹּקֶר was (and still is) written דיבור – דִּבּוּר ;שולחן – שֻׁלְחָן ;בוקר, etc., and עיוור – עִוֵּר ;הוועד – הַוַּעַד, etc. When the Academy of the Hebrew Language adopted the rules of the Va'ad ha-Lashon in 1969, this alternative was abrogated and the basic ruling alone maintained.

The resolution of the Academy was submitted to the Ministry of Education and Culture and published with the minister's signature. A committee appointed by the ministry started consultations to decide at what stage and by what didactic methods supplemented unvocalized spelling should be taught in schools.

Since that time, the Academy revised the spelling rules once again. The decisions made in 1994 can be consulted in R. Gadish (ed.), Kelalei ha-Pissuk, Kelalei ha-Ketiv Hasar ha-Nikud, Leshonenu La-Am, special issue, 4th edition (2002).

VOCABULARY

How the vocabulary of the "dead" language was adapted to the requirements of expression in all fields of life and thought is taken by many as the most outstanding achievement of the revival period. True, in less than two generations, thousands of new words and new uses of words have become part of the Hebrew lexicon. However, in this respect at least, Hebrew never was really dead; in literature and occasionally in speech, new words were being coined continually, and while these activities did not cover all domains of life, contents were not restricted to religion, philosophy, poetry and the like. Medieval literature comprises works on medicine, mathematics, astronomy, and other sciences. Matters of daily life were dealt with, for example, in the vast *responsa literature, in itineraries, etc. and new impulses of modernization further widened their scope. To adduce only two examples, one of a more comprehensive character and the other one particular word: (1) many names of animals used today are not words invented by Ben-Yehuda or after him, but first appeared around 1870 in *Mendele Mokher Seforim's "Natural History"; (2) if the Hebrew word for "passport" occurring in medieval literature – תִּיּוּר (tiyyur) or כְּתַב־תִּיּוּר (kètav tiyyur) – had not been overlooked, there would have been no need for the recent use of the ancient word דַּרְכּוֹן (darkon, a coin) in this sense.

As the last example shows, new words have been formed for concepts that had already been expressed by other words in the past. *Bialik's idea (in his essay "Ḥevlé Lašon," 1891) that total acquaintance with the store of the language ought to precede coining new words could not be followed, let alone his wish that a complete dictionary of all Hebrew writing should be the source for new word creations. In *Aḥad Ha-Am's view (put forward in his essay Ha-Lašon we-Sifrutah, 1909), the vocabulary could only be expanded by creative artists and thinkers who would be guided by the genius of the language. Bialik, though he agreed with him with regard to genuine autonomous creation, insisted upon the necessity to supply, even by designed regular activity, all the words needed, particularly those that had their semantic counterparts in other languages.

Bialik thus approved of, and later participated in, the work undertaken by Ben-Yehuda and the Va'ad ha-Lashon. While many words were, and are, to an ever-increasing degree, invented by writers and experts in their special fields, the principles and methods that guided Ben-Yehuda and his circle were followed by almost all authors of words, if not through conscious abidance, then by imitation and analogy.

The sources and ways for extending the vocabulary were expounded in Ziḵronot Waʿad ha-Lašon I (p. 7f.): (1) The best method to glean lexical items for modern use was "to search all departments of Hebrew literature and gather from them words…" If the meaning of an ancient word in its original context is doubtful, efforts should be made to clarify its interpretation; if, however, no decision can be reached, coining a new word is preferable to using a contested word. Thus, even today a biblical or later word may be submitted to the Academy for inclusion in a dictionary, but may be opposed by some members, or rejected by the majority, because commentators disagree on its meaning at its source. (2) As far as necessary, Aramaic words may be accepted, but these, unless they are already well known in their original form, are to be reshaped to fit Hebrew pattern and grammar, as happened to the Aramaic עוֹבָדָא (ʿovada) which became עֻבְדָּה (ʿuvda) ("fact"), changing both its vocalization and gender.

Ben-Yehuda's design to exploit freely the abundant vocabulary of Arabic has been accomplished to a limited extent only. Most of the words from an Arabic source found in literary and higher colloquial language are either medieval borrowings: אֹפֶק (ʿofeq), קֹטֶב (qoṭev), קֹטֶר (qoter), מֶרְכָּז (merkaz), and more, or are due to Ben-Yehuda himself: הֳגִירָה (hagira), הָצְהָרָה (haṣhara), זִבְדָּה (zivda), אָדִיב (adiv), לְטִיפָה (leṭifa), etc. Later, few new words entered standard Hebrew from Arabic, presumably because the creators of the new vocabulary came in the main from circles who knew no Arabic. When borrowing from foreign languages, they preferred the European, often international, word to the Arabic one, e.g., נֵיטְרָלִיּוּת (neṭraliyyut; "neutrality") to the suggested חִיּוּד (ḥiyyud: from the Arabic ḥiyâd). It is all the more remarkable that substandard speech has a high proportion of Arabic words and phrases.

In earlier years, borrowing from non-Semitic languages was firmly rejected with the exception of words with a Hebrew-like form or already in frequent use, but this attitude was later abandoned. Yet, many speakers of Hebrew even today frown upon words taken from European languages, such as, אֳקָדֵמְיָה (aqademya) and אוּנִיבֶרְסִיטָה (universita) proving that such words are felt to be foreign unwieldy elements and that the prior general attitude in this respect did not merely stem from the extremism of a few. On the whole, however, international technical terms are now widely adopted not only in specialized publications but in newspapers and books for the general public. Colloquial speech, too, comprises many foreign words, partly perhaps due to a passing snobbish fashion. *Even-Shoshan's seven-volume "New Dictionary" (1966–1970) contains 3,448 foreign, mainly international, words among its 33,549 basic items, and in the "Dictionary of Terms in Photography," published by the Academy in 1966, there are 53 borrowed international words among its 700 items. An important restriction on the borrowing of words of non-Semitic origin is the structure of the Hebrew verb which is formed according to severe rules, e.g., that certain vowels must appear in certain positions in verb forms; that only a limited number of consonants can constitute a verb-root, etc. As a matter of course, the necessary coining of an original Hebrew verb often also leads to the replacement of the corresponding foreign noun.

According to the principles of the Va'ad ha-Lashon, words should be created "in agreement with the rules of grammar and analogy." As far as possible, they are to be derived from roots found in biblical and talmudic literature and, in the second place, from Aramaic and other Semitic, especially Arabic, roots. To establish new scientific terms, one should aim at the essential signification, not the literal meaning, of the words of other languages. Newly coined words have not only to be grammatically correct, but pleasing to the ear and appropriate to the spirit of the language.

Contrary to the intention to avoid expressing only the literal and etymological meaning of the foreign word, loan-shifts and loan-translations have been and are an ever-growing source for new uses of existing Hebrew words. As in every language throughout the ages, Hebrew words also contract new meanings under the influence of particular applications of corresponding words in other languages, and foreign compound words and phrases are rendered in Hebrew by literally translating their components. English "crane," for the hoisting machine, French "grue," etc. have brought about the Hebrew עֳגוּרָן (ʿaguran) derived from the name of the bird; and German "Kindergarten" has engendered גַּן־יְלָדִים (gan-yeladim; "garden" (of) "children"). Nowadays, the principle of such semantic borrowings is seldom debated; only innovations that are felt to be too farfetched and removed from prevailing usage are rejected. In word formation, modern Hebrew, for the most part, follows the methods inherited from former stages of the language. The available noun and verb patterns are used to the full for innovations. Yet, some possibilities of derivation and combination that in older Hebrew were realized in relatively small measure are now put to use more extensively and, as some maintain, even excessively. The following deserve special mention and exemplification:

(1) Many nouns and adjectives are derived from noun bases by adding suffixes:

(a) -an, fem. -anit for nouns, as in: חֳצוֹצְרָן (ḥaṣṣeran, "trumpeter"); תּוֹתָחָן (totaḥan, "artilleryman"); דּוֹדָן (dodan), fem. דּוֹדָנִית (dodanit, "cousin"); מַהְפְּכָן (mahpěḵan, "revolutionary," n.); תִּיקָן (tiqan, "cockroach,"), from תִּיק (tiq, "envelope," i.e., the protective shell of the insect's eggs).

(b) -ay, fem. -aʿit for nouns, as in: עִתּוֹנַאי (ʿittonay, fem. עִתּוֹנָאִית ʿittonaʾit, "journalist"); בּוּלַאי (bulay, "philatelist"); אֳוִירַאי (awiray, "airman"); מְכוֹנַאי (měḵonay, "machinist"); טֵלֵפוֹנַאי (ṭelefonay, fem. טֵלֵפוֹנָאית (ṭelefonaʿit), "telephone operator"); סְטָטִיסְטִּקַאי (staṭisṭiqay, "statistician").

(c) -on, fem. -ónet, often for diminutive nouns: דֻּבּוֹן (dubbon, "young bear, teddy bear"); יַלְדּוֹן (yaldon, "small boy"), יַלְדֹּנֶת (yaldonet, "small girl"); שֵׁדוֹן (šedon, "sprite").

(d) -i, fem. -it, mainly for adjectives: צְבָאִי (ṣevaʾi, "military"); גַּמָּדִי (gammadi, "dwarfish"); אַפְסִי (afsi, "amounting to nothing"); תְּהומִי (těhomi, "abysmal"), עֳנָקִי (ʿanaqi, "colossal"). The suffix -i is also widely used to derive adjectives from compounded pairs of nouns, as in צְפוֹן־מִזְרָחִי (ṣěfon-mizraḥi, "north-eastern"), from צְפוֹן־מִזְרָח (sěfon-mizraḥ, northeast); גַּב־לְשׁוֹנִי (gav-lěšoni, "dorsal," in phonetics), from גַּב־לָשׁוֹן (gavlašon, "dorsal surface of the tongue"); כְּלַל־אֱנוֹשִי (kělal-ěnoši, "all-human, universal"). This mode of derivation, found in the Bible in gentilitial names, like בֶּן־יְמִינִי (Ben-Yěmini, "Benjaminite") and בֵּית־הַלַּחְמִי (Bet-Hallaḥmi, "Bethlehemite"), has also been extended to compounds whose first member is a quantifier, as in חַד־כִּוּוּנִי (ḥad-kiwwuni, "unidirectional"); דּוּ־לְשׁוֹנִי (du-lěšoni, "bilingual"); רַב־צְדָדִי (rav-ṣědadi, "many-sided"); or a preposition, as in בֵּין־לְאֻמִּי (ben-lěʾummi, "international"); קְדַם־מִקְצוֹעִי (qědam-miqṣoʿi, "pre-professional"): עַל־אנוֹשִׁי (ʿalěnoši, "superhuman").

(e) -it, for nouns, some diminutive (besides being the feminine form of -i): מְכוֹנִית (měḵonit, "automobile"); מוֹנִית (monit, "taxi"); יָדִית (yadit, "handle"); מַפִּית (mappit, "napkin"); תָּוִית (tawit, "label").

(f) -ut, for abstract or collective nouns: בּוֹרְרוּת (borěrut, "arbitration"): צִּיּוֹנוּת (ṣiyyonut, "Zionism"); רוֹקְחוּת (roqẹḥut, "pharmacology"); מְיַלְּדוּת (meyallědut, "obstetrics"); עִתּוֹנוּת (ʿittonut, "press").

(g) Several of the foregoing suffixes may combine to form new derivations, such as: מַהַפְּכָנִי (mahpěḵani, "revolutionary," adj.); מַהְפְּכָנוּת (mahpěḵanut, "revolutionism"); תּוֹתְחָנוּת (totěḥanut, "artillery"); עִתּוֹנָאוּת (ʿittonaʾut, "journalism"); גַּמָּדִיּוּת (gammadiyyut, "dwarfishness"); אַפְסִיּוּת (afsiyyut, "worthlessness").

(2) New nouns are built by joining elements of two other words, particularly when this is suggested or facilitated by both words having one or more consonants in common or by the second word beginning with a glottal stop (ʾalef) which can easily be omitted. קוֹלְנוֹעַ (qolnoaʿ, "cinema") is but a simple joining of קוֹל (qol, "sound") and נוֹעַ (noaʿ, "movement"), while אוֹפַנּוֹעַ (ʾofannoaʿ, "motorcycle") joins אוֹפַן (ʾofan, "wheel") and נוֹעַ (noaʿ). Two original consonants are omitted in דַּחְפּוֹר (daḥpor), a blending of the verbal roots דחף (d.ḥ.f.) and חפר (ḥ.f (p).r); with the recurring pair ח () and פ (f (p)) inserted only once, the sequence דחפר (d.ḥ.f (p).r) is left and shaped into a noun with the vowel sequence a.o frequent in nouns. On the same vowel pattern רַמְזוֹר (ramzor, "traffic light") is formed from the verbal root רמז (r.m.z, "to indicate") and the noun אוֹר (ʾor, "light") whose initial א 'alef is elided. The popular creation שְׁמַרְטַף (šěmarṭaf, "babysitter") is compounded from שמר (š.m.r., "to watch") and טַף (ṭaf, "children"), but the Academy prefers שׁוֹמֵר-טַף (šomer-ṭaf) modeled after the biblical שׁוֹמֵר־סַף (šomer-saf, "keeper of the door").

(3) Among verbal innovations the amount of denominative verbs is significant: רִשֵּׁת (riššet, "to cover with a net") comes from רֶשֶׁת (réšet, "net"); קִרְקַע (qirqʿa, "to ground [an aircraft]") from קַרְקַע (qarqaʿ, "ground"): נִתֵּב (nittev, "to pilot") from נָתיב (nativ, "path"), and numerous others, especially scientific, technological, and military terminology. For such new active verbs, the pattern piʿẹl is preferred with hif ʿil left far behind and paʿal (qal) almost entirely neglected.

(4) Many of these new denominative verbs are derived from nouns with prefixed or suffixed formatives. Thereby, new roots, mostly quadriliteral, have entered the language: מִרְכֵּז (mirkez, "to centralize"), with it the passive participle מְמֻרְכָּז (měmurkaz), and the action noun מִרְכּוּז (mirkuz) have been derived from מֶרְכָּז (merkaz, "center") to differentiate from the former verb רִכֵּז (rikkez, "to concentrate") which shows the original root רכז (r.k.z) מִסְפֵּר (misper, "to number") contains in its secondary root מספּר (m.s.p.r) the consonants of מִסְפָּר (mispar, "number"), a noun derived from the primary root ספר (s.f (p).r). The relation between תִּזְמֵר (tizmer, "to orchestrate"), תִּזְמֹרֶת (tizmóret, "orchestra"), and the primary root זמר (z.m.r) is similar. In a piyyuṭ by Eleazar *Kallir (of the early Middle Ages) there is the verb הִתְמִיר (hitmir) originating from תְּמוּרָה (těmura, "change") which in turn is based on the primary root מור (m.w.r); the verb הִתְמִיר (hitmir) has now passed from its remote literary source into modern use in the meaning "to substitute" in chemistry. From the primary root חמץ (ḥ.m.ẓ) Ben-Yehuda formed the noun חַמְצָן (ḥamẓan, "oxygen"), and this served as a base for the new verb חִמְצֵן (ḥimṣen, "to oxydize").

(5) Another way to form denominative verbs is to derive new roots from contractions or acrostics of compound words. Thus, from דְּין־וְחֶשְׁבּוֹן (din-wě-ḥešbon, "account, report") first the acrostic דּוּ״חַ (duaḥ, "report") came into use, and then the verb דִּוַּח (diwwaḥ, "to report") was formed with the artificial root דוח (d.w.ḥ.). In order to obtain a Hebrew verb for "to internationalize," to which בֵּין־לְאֻמִּי (ben-lěʾummi, "international") did not lend itself, a contracted root בנאם(b.n.ʾ.m) had to be presumed to arrive at the desired verb בִּנְאֵם (binʾem) and its action noun בִּנְאוּם (binʾum, "internationalization"). However, this presumption is not so farfetched, since there is a Hebrew noun אֻמָּה (umma), besides לְאֹם (ʾom) for "nation."

(6) In analogy to several verbs of the šaf ʿel formation inherited from biblical and later Aramaic and Hebrew, some new causative verbs and action nouns with the prefixed š- have been created from existing roots, mainly where other verb formations had already been exploited for the same root. To these innovations, some of which have been sanctioned by the Academy, belong שִׁחְזֵר (šiḥzer, "to restore"), root חזר (ḥ-z-r, "to return"), and its action noun שִׁחְזוּר (šiḥzur); שִׁקֵּם (šiqqem, "to rehabilitate") with שִׁקּוּם (šiqqum) as action noun, root קום (qw-m, "to rise"); שִׁפְרֵט (šifreṭ, "to elaborate") derived from פְּרָט (pěraṭ, "detail"); שִׁכְפֵּל (šiḵpel, "to duplicate, multiply (written matter)"), etc. Among the first and most widely used of these new words were שִׁחְזוּר (šiḥzur) with the meaning of restoration of a previous condition inherent in its root חזר (ḥ-z-r), and שִׁקּוּם (šiqqum), which intrinsically only means "causing to rise, erecting," but was used in contexts entailing the connotation of "again." Many speakers, therefore, came to attribute this meaning of remaking or redoing to the šaf ʿẹl formation, and by way of vindicating this semantic shift, some even interpreted the prefixed š- as an abbreviated שׁוּב (šuv, "again"). On this assumption, more verbs and action nouns with initial š-, corresponding to English re-, have been formed and in part accepted: שִׁחְלֵף (šiḥlef, "to re-exchange"), root חלף (ḥ-l-f) – in another sense שַׁלְחֵף < שַׁחְלֵף (šalḥef < šaḥlef) is already found in the Aramaic of the Targum and Talmud; שִׁעֲרוּךְ (šiʿiaruḵ, "reassessment"), root ערךְ (ʿ-r-k); שִׁזְרַע (šizraʿ, "to resow"), root זרע (z-r-ʿ), שִׁגְזוּר (šigzur, "back formation" in linguistics), root גזר (g-z-r- "to derive"), etc.

(7) A considerable number of passive verbal adjectives has been adopted with the vowel sequence a-i inserted in the root and corresponding in meaning to French and English adjectives in -able, -ible. The first of these probably was שָׁבִיר (šavir, "breakable"), followed by קָרִיא, (qari, "readable," "legible"), סָביר (savir, "reasonable"), כָּבִיס (kavis, "washable"), חָדִיר (ḥadir, "penetrable"), דָּחִיס (daḥis, "compressible"), and more. However, this pattern has at all times served in the formation of other adjectives (as the biblical ʿašir "rich") and of nouns (as the biblical qaṣir "harvest"). Its application to defective roots meets with difficulties; its use is limited to derivations from paʿal (qal) verbs, and its corresponding abstract noun is ambiguous (e.g., דְּחִיסוּת děḥisut may be understood as "compressibility" and as "[state of] compression," from dahus, "compressed"). Words of this semantic category are, therefore, also formed in other ways, either with the suffix -i appended to an action noun, as in שִׁמּוּשִׁי (šimmuši, "practical"), from שִׁמּוּשׁ (šimmuš, "practice, use"), or, as in classical Hebrew, either by the use of passive participles, such as, מִתְקַפֵּל (mitqappel, "collapsible," "folding"); נֶאֶכַל (neʾěḵal, "edible"); מִטַּלְטֵל (miṭṭalṭel, "portable"), etc., or by compounding בֶּן־ ben- or בַּר־ bar- with abstract nouns, mostly action nouns, as in בֶּן־סֶמֶךְ (ben semeḵ; "reliable"), בֶּן־בּוּז (ben-buẓ [Bialik], "contemptible"); בַּר־בִּצּוּעַ (bar-biṣṣuaʿʿ "executable"); בַּר־בִּטּוּל bar-biṭṭul, "abolishable"), etc.

GRAMMAR

In 1905, the teachers' convention agreed to Yellin's proposal for a standardized orthography based on the biblical vocalization system. This, to a large extent, led to the acceptance of biblical Hebrew grammar for modern Hebrew. The spelling and vocalization adopted determined the form of words and their inflection, though in this sphere, too, usage had to be normalized to eliminate variations and prosodic peculiarities of the Bible text.

In 1910, this topic was discussed in a meeting of the Va'ad ha-Lashon in which Eliezer Ben-Yehuda, Yisrael Eitan, David Yellin, Aharon Masie, and Yosef Meyuḥas took part. The summary of this debate, published in Ziḵronot Waʿad ha-Lašon II (2nd ed., 1929, p. 17ff.) has lost little of its import; many of the arguments set forth then are still heard whenever there is doubt about the preferable form and inflection of a word or a class of words. The discussion then originated from one particular question, i.e., the correct plural for the mishnaic אֹגֶן (ʾogen, "rim, brim"), whether it should be אֳגָנִים (oʾganim) according to grammar based on the Bible, or (אָגְנִים אוגנים) (ʾognim) as found in mishnaic Hebrew. Many aspects of modern grammar were treated in the light of ancient literary sources during this debate.

In conclusion, the following resolution was proposed by D. Yellin and adopted: "We take from talmudic and midrashic literature words and expressions which we need and new grammatical forms supplementing those found in the Bible. Talmudic words accepted are to be given a Hebrew form whenever possible. For verbs no new forms are needed if in the Bible there are corresponding forms." The significant words in this resolution provide that neither in vocabulary nor in grammar should anything available in the Bible be replaced by elements from later literature. Regarding the special problem of the plural of nouns, such as אֹגֶן (ʾogen), both forms were admitted, but later the "Dictionary of Technical Terms," published in 1929 by the Va'ad with H.N. Bialik as one of its editors, contained only the plural form אֳגָנִים (ʾoganim) in conformity with biblical grammar.

The principles adopted by the Va'ad ha-Lashon were observed, with few exceptions, by teachers and in textbooks. According to these norms, the teaching of grammar, on the whole, not only excluded the divergent traditions of Hebrew and the innovations found in post-biblical literature, but also disregarded the language in which Hebrew literature had been written since the end of the Haskalah. It was to the principles of the Haskalah that normative grammarians now reverted. The literature before the revival of Hebrew speech in Palestine, especially since Mendele Mokher Seforim, had not submitted to the restrictions imposed by the Haskalah, but had freely blended biblical elements with talmudic and later grammatical forms as well as with words and phrases from all periods and even borrowings from modern European languages. However, as modern literature and speech have continued to grow, grammar based on the Bible has proved inadequate for all the new material.

The strict adherence to what was known and held in biblical grammar is well exemplified by A.Y. Shapiro in his More Nevuḵé ha-Lašon (Warsaw, 1909). The author corrects about 140 words and grammatical forms found in the writings and speech of his contemporaries, naturally according to his views on the Bible text and to the conclusions he draws from it. Thus, he rejects נֶאֶבַד for הִתְאַנַּח ;אָבַד for אֳרוּכָה ;נֶאֱנַח for אֳרֻכָּה (fem. of אָרֹךְ, "long"), the infinitives לֵילֵךְ ,לֵישֵׁב ,לֵידַע, etc., for the biblical forms לָלֶכֶת ,לָשָׁבֶת ,לָדַעַת, etc.; infinitives such as לִקְרוֹת for יֹפִי ;לִקְרֹא both in the absolute and construct case for יְפִי; the imperfect אֶלְמוֹד for עָמוּס; אֶלְמַד for עוֹמֵס ("carrying, burdened"): …הֶרְאָה אֶת… לְ for …הֶרְאָה אֶת… אֶת; etc. Although the author, in an appendix, shows that some of these and other non-biblical forms are found in talmudic literature, he does not approve of their use in modern language. However, these forms and many more have in fact existed in the literature, or in certain traditions of Hebrew, for centuries and are accepted by some of the best modern writers, their selection being but a matter of personal style.

As Hebrew is the paramount unifying factor of modern national culture in Israel, a distinction had of necessity to be made between the standard language taught in schools and used in public addresses, broadcasting, and the like, and the individual idiom of creative writers and the traditions of the various Jewish communities in reading religious and other texts that naturally also influence their everyday speech. The work of establishing a normative grammar of modern Hebrew – one of the chief tasks of the Va'ad ha-Lashon and the Academy of the Hebrew Language – will understandably take a long time and, in fact, imposes itself constantly anew in response to cultural developments. However, once the foundations have been laid and become general usage, much can be left to natural growth without interference of any linguistic authority. Even today the greater part of new words and word forms used spontaneously by individuals already conforms to grammar rules.

The new, Hebrew grammar is gradually being built by two separate activities: by comprehensive discussion and decision on systematic divisions of grammar and by ad hoc instructions on particular problems submitted by writers, teachers, journalists, and other members of the public or arising from the work of terminological committees. The former course is naturally preferred, yet it is lengthy and cannot answer urgent needs; therefore, the latter is unavoidable, although its ad hoc directives have occasionally to be amended to agree with a subsequent comprehensive ruling.

The foremost aim of the Va'ad ha-Lashon in its systematic treatment of grammar was to decide on words for which the biblical text does not provide sufficient exemplary evidence or offers several divergent forms of one word, e.g., for the noun לְטָאָה ("lizard"), found only in this form once in Leviticus, two forms with pronominal suffixes may be inferred: לְטָאָתִי or לִטְאָתִי; by analogy, two forms for the post-biblical מְחָאָה ("protest") would be possible; אֲגַם ("pool," "pond") has two inflected forms belonging to different paradigms: אֲגַמִּים and אַגְמֵיהֶם, both in Exodus. Whenever there was no doubt about the biblical form of a word, the Va'ad accepted this precept and allowed only very few exceptions dictated by firmly established usage, e.g., permitting כִּתְבֵי הַקֹּדֶש ("Holy Scriptures") besides the form based on the Bible כְּתָבֵי הַקֹּדֶשׁ from כְתָב with an unchangeable qameṣ. This principle inevitably led to a twofold treatment of words of one and the same morphological pattern: Words taken from the Bible went one way, and those coming from later sources or coined recently went another. Thus, the rule for nouns of the pattern qěṭal to which כְּתָב belongs, lists the biblical words whose qameṣ is to be unchangeable, and provides for the change of qameṣ to pattaḥ or šewa only in words from later sources, such as שְׁטָר ("writ, note") – שְׁטַר־חוֹב ("note of debt," sing. construct state) – שִׁטְרֵי־חוֹב (plur. construct state).

When the Academy continued this work of the Va'ad ha-Lashon, the renewed debate led to a fundamental change. Now, the rules are to deal with modern Hebrew as a whole, and the dichotomy of its vocabulary by reason of its sources, whether biblical or post-biblical, has been abandoned. It is no longer a matter of course that for each biblical word its biblical inflection be accepted in the modern language. If this is to be done, and, in general, it is, the issue is open to discussion and subject to decision in accordance with the tendency to allow well-established traditions and usages their proper place, and to make each new rule as comprehensive as possible. Most of the rules still have their exceptions, of course, but these are few, and they sometimes include biblical forms or state their existence without recommending their use.

So far, only the rules for the inflection of nouns have been systematically discussed and partly established. The arrangement of the rules follows the alteration of the vowels in each class of nouns, this being the prominent feature in Hebrew inflection. Each rule is the outcome of a thorough examination of the ways in which the various sources of the language treated the vowels in inflection. The rules for qameṣ gadol and pattaḥ have been published (Ziḵronot 7–8, 1962, p. 91ff and 13, 1967, p. 7f.), and the rules for the other vowels were decided upon and issued by the Academy in later sessions, the most recent publication being in Leshonenu La-Am, 51–52 (2000–1), pp. 153–98.

As an example, paragraph 8 of section 2 in chapter I will be given here with some added remarks:

The qameṣ gadol is stable in the endings ־ָר ,-תָן ,־ָן in nouns denoting occupations and qualities, such as םֶכיֵנדמַל – ןָדְמַל; יֵנ ָל ְּב ַק – ןָל ְּבק; יֵנתְואַּג – ןָתְואַּג; םֶכירָל ְבַל – רָל ְבַל; רָמיִּב – יֵרָמיִּב;רָטוטרַמְס – יֵרָטוטרַמְס; רָלדְנ ַס – םֶכירָלדְנ ַס.

The qameṣ is stable in the nouns ןָתי ֵא – יֵנתי ֵא, ןָמיִס – יֵנ ָמי ִס, and in loanwords, such as ןָמֹור, ן ָמ ֹופרְּג, etc.

In other nouns, the qameṣ changes in inflection: – תֹונ ֲח ְלשןָחְלֻש (constr.); ןָחְלֻּפ – יֵנ ֲח ְלפֻּ; ןָּדְמֻא – יֵנדְמֻא; ןָיְנע – ם ֶכי ֵנ ְי ְנע; ןָיְנ ִּב – יֵנְיְנ ִּב; רָּבְכע – יֵרְּבְכע; ןָּברָק – יַתֹונ ְּברָק; רָסְפט – יֵרְסְפט; רָדוס – יֵרְדוס; רָלֹוק – םֶכירְלֹוק.

This paragraph presupposes paragraph 1 of section 1 which provides that "every qameṣ gadol, occurring in the absolute state in a stressed syllable, changes to pattaḥ in the singular construct state and before the pronominal suffixes םֶכ-, ןֶכ-." Therefore, ןָחְלֻש, e.g., in these two contexts becomes ןַחְלֻש and םֶכְנחְלש respectively.

Of the 21 words adduced as examples in this rule, only seven are biblical: ןָתי ֵא, ןָחְלֻש, ןָּברָק, ןָיְנ ִּב, ןָיְנע, רָּבְכע, רָסְפט. (The last word, of Sumerian-Accadian origin, occurs twice in the Bible, once with ḥireq and once with pattaḥ in the first syllable.) Another nine words, partly Greek or Latin borrowings, are found in talmudic-midrashic literature: רָלדְנ ַס, רָלֹוק, רָדוס, ןָל ְּבק, ןָתְו ַאַּג, ןָחְלֻּפ, ןָּדְמֻא, ןָמיִס, רָל ְבַל, and one, ןָדְמַל is found in medieval writings (and also in Yiddish), but there are many newly created words of the same formation. Two words are modern derivations from older ones: רָטוטרַמְס ("rag picker") from the talmudic טוטרַמְס ("rag") and רָמיִּב ("stage technician") from the originally Greek ה ָמי ִּב ("stage"). The remaining two, ןָמֹופרְּג, ןָמֹור, are contemporary loans from European languages.

The salient point here is that, without regard to their history, all these words are integrated in the modern vocabulary and divided with respect to their inflection not necessarily in conformity to biblical grammar. The fact that they do not behave uniformly in inflection has historical reasons. The group with changing qameṣ follows three of its members – ןָחְלֻש, ןָּברָק, רָּבְכע – for which the Bible text has šewa, or šěwa compositum, replacing qameṣ in the relevant inflected forms. The other group, with stable qameṣ, complies with the usual pronunciation of most of its members.

Outside of the systematic treatment of grammatical and other problems by the Academy, ad hoc solutions of specific questions deal not only with morphology, but with syntax and style as well. A few examples must suffice here. In the field of morphology it is often necessary to fix the exact spelling and vocalization of old words that have been handed down in several forms. The vowels of the talmudic noun ןפוג ("character of script") are uncertain; thus of the forms ןָפֹוג, ןָפוג, ןֶפֹג which are found, the first has been chosen. Even for the verb עיזה ("to sweat," in the hif ʿil, two vocalizations have been in use: עי ִז ֵה and עי ִז ִה; the choice fell on עי ִז ֵה because it agrees with the root suggested by the inflection of the biblical noun הָעֵז ("sweat") which alone is in common use today (not עַזֶי). Committees on terminology, when proposing a new word, are often in doubt about its grammatical form. Thus, לָכמ ("container") has been selected instead of לַכמ previously chosen by the Va'ad ha-Lashon. Foreign words admitted into the language require their Hebrew plural to be determined. Thus for the plural of םומי ִס ְק ַמ ("maximum") the form ת ֹוא ָמי ִס ְק ַמ has been proposed in the same way as mishnaic Hebrew dealt with similar Greek and Latin nouns (ן ֹולי ִו – Latin "vellum" – plural ת ֹוא ָלי ִו).

Syntactic structure in translated literature and in journalistic writing has been greatly influenced by European languages (now mainly English). One of the results, for example, is the frequent appearance of non-restrictive, continuative relative clauses, such as, תִי ַּב ַה ךֹות ְל טַלמִּנ ְׂש בָּנגַה יֵרֲחַא ופ ְדר םירְטֹושַה ב ֹורקַה. ("The police pursued the thief, who escaped into the nearest house"). Although this use is found neither in the colloquial language nor in that of writers whose Hebrew is considered exemplary, it is frequent in journalese and officialese. Some linguists do not condemn it on this level of the language, and the same applies to other syntactic structures, equally foreign to more elevated and conservative style.

Modern Hebrew as a Semitic language, with an ancient literary heritage still cherished and studied, was already exposed to the impact of the modern world and of modern non-Semitic languages when it only was the vehicle of literary revival and before it became a fully living language. Whoever took part in the revival of the language, in writing or in speech, was aware of this position, its requirements and consequences. But for the past 80 years at least, Ben-Yehuda and his collaborators and their successors have made a conscious effort to develop Hebrew and adapt it to modern use on the lines on which, in their view, it would have developed if its natural life had continued without interruption into the 20th century. In fact many other languages which have not passed a period of suspended animation now face problems quite similar to those of modern Hebrew. What Hebrew experienced now has happened to it before, for example, in the talmudic period and in the later Middle Ages, when not only new words were formed or borrowed and old words were used to refer to new objects, but the morphological, syntactic, and conceptual structure of the language changed in part, both by direct imitation of other languages and under the influence of their manner to organize the relation between words and concepts.

One of the characteristics of modern Hebrew is the speed of the changes in all respects. Thus it offers much interesting material to the linguist to show the trends of its evolution and to discover general linguistic facts and processes in it. Two phenomena: "Westernization" and "re-Hebraization" (much discussed in treatises on language policy, especially by Rosén, Ben-Hayyim, and Bendavid), in the recent development of the language are obvious to all observers. The "ancient language being in a new reality" absorbs concepts and forms of Western languages through cultural contacts, through more or less apt translation, immigration, and bilinguism, etc. The wish to strengthen the inherited Hebrew component is obvious and may be realized through extended Hebrew education, more intense study of classical writings, the growth of modern literature imbued with old language tradition and the increased number of its readers, competent guidance of language development and by adapting old forms to modern contents.

[Eli Eytan]

BIBLIOGRAPHY:

PRE-BIBLICAL GENERAL: C. Brockelmann, in: Handbuch der Orientalistik, vol. 3, Semitistik, Abschnitt 1 (1953), 40–58 (Canaanite dialects and Ugaritic), 59–70 (Hebrew); G. Garbini, Il Semitico di Nord-ovest (1961): Gelb, in: Journal of Cuneiform Studies, 15 (1961), 27–47; A. Goetze, in: Language, 17 (1941), 127–38; Greenberg, in: JAOS, 72 (1952), 1–9; Z.S. Harris, Development of the Canaanite Dialects (1939); Moran, in: Wright, Bible, 59–85; Polotsky, in: A World History of the Jewish People 1st series, 1 (1964), 104–11; Roessler, in: ZDMG, 100 (1950), 461–514; Von Soden, in: WZKM, 56 (1960), 177–91. EGYPTIAN MATERIAL: W.F. Albright, The Vocalization of the Egyptian Syllabic Orthography (1934); idem, in: JPOS, 8 (1928), 223–56; idem, in: JAOS, 74 (1954), 222–33; Borée, M. Burchardt, Die altkanaanaeischen Fremdwoerter und Eigennamen im Aegyptischen (1909–10); Moran, in: Orientalia, 26 (1957), 339–45; K. Sethe, Die Aechtung feindlicher Fuersten, Voelker und Dinge auf altaegyptischen Tongefaesscherben des mittleren Reiches (1926); J. Simons, Handbook for the Study of Egyptian Topographical Lists Relating to Western Asia (1937); G. Posener, Princes et Pays d'Asie et de Nubie (1940). CUNEIFORM MATERIAL FROM TAANACH: Albright, in: BASOR, 94 (1944), 12–27; Gross, ibid., 190 (1968), 41–46; Hillers, ibid., 173 (1964), 45–50: F. Hrozny, in: E. Sellin, Tell Taʿannek, Bericht ueber eine… Ausgrabung in Palaestina (1904), 113–22. CUNEIFORM MATERIAL FROM EL-AMARNA: Albright, in: BASOR, 89 (1943), 7–17; F.M.T.Boehl, Die Sprache der Amarnabriefe (1909); E. Dhorme, in: RB, 22 (1913),369–93; 23 (1914), 37–59, 344–72; Ebeling, in: Beitraege zur Assyriologie und semitischen Sprachwissenschaft, vol. 8, pp. 39–70; J.A. Knudtzon, Die El-Amarna Tafeln (1915); Moran, in: Orientalia, 29 (1960), 1–19; idem, in: Journal Of Cuneiform Studies, 4 (1950), 169–72; A.F. Rainey, El-Amarna Tablets 359–379, suppl. to J.A. Knudtzon's Die El-Amarna Tafeln (1970); For primary references see: R. Borger, Handbuch der Keilschriftliteratur, 1 (1967), 237f. AMORITE CUNEIFORM MATERIAL: G. Buccellati, The Amorites of the Ur III Period (1966): T. Bauer, Die Ostkanaanaeer (1926); A. Finet, L'Accadien des Lettres de Mari (1956); E. Dhorme, in: RB, 37 (1928), 63–79, 161–180; 39 (1930), 161–78; 40 (1931), 161–84; Gelb, in: Atti della Accademia Nazionale dei Lincei, Rendiconti, Classe di scienze morali, storiche e filologiche, serie 8, 13 (1958), 143–64; idem, in: JAOS, 88 (1968), 39–47; H.B. Huffmon, Amorite Personal Names in the Mari Texts (1965). ADD. BIBLIOGRAPHY: ONOMASTICS: S.C. Layton, Archaic Features of Canaanite Personal Names in the Hebrew Bible (1990); Y. Muchiki, Egyptian Proper Names and Loanwords in North-West Semitic (1999). EL-AMARNA: D. Sivan, Grammatical Analysis and Glossary of the North-West Semitic Vocables in Akkadian Texts of the 15th13th C.B.C. from Canaan and Syria (1984); A.F. Rainey, Cannanite in the Amarna Tablets, 4 vols. (1996); W.L. Moran, The Amarna Letters (1992); idem, Amarna Studies: Collected Writings (2003). AMORITE: I.J. Gelb, Computer-aided Analysis of Amorite (1980); C.H. Gordon, "Amorite and Eblaite," in R. Hetzron (ed.), The Semitic Languages (1997), 1001–13; M.P. Streck, Die Amurriter, die onomastische Forschung, Orthographie und Phonologie, Nominalmorphologie (2000). BIBLICAL. The best grammar is still that of G. Bergstraesser, Hebraeische Grammatik (1918–29), which, though a torso, is an impressive piece of scholarship in the field of phonetics and the verb. An excellent short account is contained in G. Bergstraesser's Einfuehrung in die semitischen Sprachen (1928, repr. 1963), 36–46. For the development of stress see M. Lambert, REJ (1890), 73–77; J. Cantineau, in: Bulletin d'Études Orientales de l'Institut Français de Damas (1931), 81–98. Important grammars are: H. Bauer and P. Leander, Historische Grammatik der hebraeischen Sprache des Alten Testaments (1922; does not contain syntax); W. Gesenius, E. Kautzsch and A.E. Cowley, Hebrew Grammar (19132 and many reprints; translated from the 28th German edition of 1909); P. Joüon, Grammaire de l'Hébreu Biblique (1923). Important material is contained in E. Koenig, Historisch-kritisches Lehrgebaeude der hebraeischen Sprache (1881–97). Biblical Hebrew against its general Semitic background is described in C. Brockelmann, Grundriss der vergleichenden Grammatik der semistichen Sprachen (1907–13); yet his Hebraeische Syntax (1956), mostly taken from the masterly syntax contained in the second volume of his Grundriss, is less good, since it lacks the general Semitic background. Good is the syntax of A.B. Davidson, Introductory Hebrew Grammar; Hebrew Syntax (19123), and even today one will often consult the masterly syntax contained in H. Ewald, Ausfuehrliches Lehrbuch der hebraeischen Sprache des Alten Bundes (18708) which was translated into English by James Kennedy as Syntax of the Hebrew Language of the Old Testament (1879). The new work of (G. Beer-) R. Meyer, Hebraeische Grammatik (1952–552) is only important because of the new material adduced; the same applies to A. Sperber, A Historical Grammar of Biblical Hebrew (1966), where, e.g., material of Jerome's transcriptions is to be found. For the language of the Qumran scrolls see E.Y. Kutscher, Ha-Lašon wě-ha-Reqaʿ ha-Lěšoni šel Mě̄gillat Yěšaʿyahu ha-Šěléma (1959; with English summary). For Hebrew inscriptions cf. S. Moscati, Stato e problemi dell'epigrafia ebraica antica (1952) where additional literature is adduced; also Aharoni and Amiran, in: IEF, 14 (1964), 138–43; Aharoni, in: IEF, 16 (1966), 1–7. For transcriptions of Palestinian place names see W. Borée, Die alten Ortsnamen Palaestinas (1930); for transcriptions of the Septuagint, G. Lisowsky, Die Transkription der hebraeischen Eigennamen des Pentateuch in der Septuaginta (1940); for those of Origen, E. Bronno, Studien ueber hebraeische Morphologie und Vokalismus auf Grund der Mercatischen Fragmente der zweiten Kolumne der Hexapla des Origines (1943). For the transcription of Arabic names in Nessana see the index of Arabic names by F.E. Day, in: C.J. Kraer (Jr.), Excavations of Nessana, 3 (1958), 352–5. For the Bible text see D. Ginsburg, Tora Něviʾim Kětuvim (19262); R. Kittel and P. Kahle, Biblia Hebraica (196213). One sample of the edition of the Hebrew University Bible Project has also appeared: M.H. Goshen-Gottstein, The Book of Isaiah, sample edition (1965). For an example of P. Kahle's views on the work of the masoretes see his Cairo Geniza (19592). The best biblical dictionaries are still F. Brown, S.R. Driver, and A. Briggs, A Hebrew and English Lexicon of the Old Testament (1906, etc.); W. Gesenius and F. Buhl, Hebraeischesund aramaeisches Handwoerterbuch ueber das Alte Testament (191516); W. Gesenius, Hebrew and Chaldee Lexicon to the Old Testament (1857, repr. 1957); and L. Koehler and W. Baumgartner, Lexicon in Veteris Testamenti libros (1953, Supplement 1958) – only the Aramaic part is up to the high standards of its predecessors. A new edition of this dictionary by W. Baumgartner, B. Hartmann and E.Y. Kutscher is appearing: Hebraeisches und Aramaeisches Lexikon zum Alten Testament, 1 (1967). For further literature see Steinschneider, Handbuch. ADD. BIBLIOGRAPHY: GRAMMAR: R.J. Williams, Hebrew Syntax: An Outline (19762); J. Blau, A Grammar of Biblical Hebrew (1976); idem, Studies on Biblical Hebrew (1995); idem, Studies in Hebrew Linguistics (1996); idem, Topics in Hebrew and Semitic Linguistics (1998); B.K. Waltke & M. O'Connor, An Introduction to Biblical Hebrew Syntax (1990); P. Joüon and T. Muraoka, A Grammar of Biblical Hebrew (1991), a translation and updating of Joüon's 1923 French grammar; J. Hoftijzer, The Function and Use of the Imperfect Forms with Nun Paragogicum in Classical Hebrew (1985); T. Muraoka, Emphatic Words and Structures in Biblical Hebrew (1985); J.C.L. Gibson, Davidson's Introductory Hebrew GrammarSyntax (1994); S.E. Fassberg, Studies in Biblical Syntax (1994); I. Ben-David, Contextual and Pausal Forms in Biblical Hebrew (1995); C.L. Miller, The Representation of Speech in Biblical Hebrew Narrative (1996); idem (ed.), The Verbless Clause in Biblical Hebrew (1999); T. Zewi, A Syntactical Study of Verbal Forms Affixed byn(n) Endings (1999). DICTIONARIES AND LEXICONS: D.J.A. Clines et al., The Dictionary of Classical Hebrew (1993– ); L. Koehler and W. Baumgartner, The Hebrew and Aramaic Lexicon of the Old Testament (2001); P. Mankowski, Akkadian Loanwords in Biblical Hebrew (2000). VERBAL SYSTEM: L. McFall, The Enigma of the Hebrew Verbal System (1982); E.J. Revell, "The System of the Verb in Standard Biblical Prose," in: HUCA, 60 (1989), 1–37; G. Hatav, The Semantics of Aspect and Modality: Evidence from English and Biblical Hebrew (1997); W. Randall Garr, "Driver's Treatise and the Study of Hebrew: Then and Now," in: S.R. Driver, A Treatise on the Use of the Tenses in Hebrew and Some Other Syntactical Questions (1998), xviii–lxxxvi; P. Gentry, "The System of the Finite Verb in Classical Biblical Hebrew," in: HS 39 (1998), 7–39; Z. Zevit, The Anterior Construction in Biblical Hebrew (1998); J. Cook, "The Hebrew Verb: A Grammaticalization Approach," in: ZAH, 14 (2001), 117–43. LATE BIBLICAL HEBREW: A. Hurvitz, The Transition Period in Biblical Hebrew (1972); R. Polzin, Late Biblical Hebrew: Toward an Historical Typology of Biblical Hebrew Prose (1976). TRANSCRIPTIONS: G. Janssens, Studies in Hebrew Historical Linguistics Based on Origen's Secunda (1982);DEAD SEA SCROLLS. General: Up to 1965: C. Burchard, Bibliographie zu den Handschriften vom Toten Meer, 2 vols. (1957–65). 1965 TO 1970: Bibliographies in: Revue de Qumran, 1–6 (1958–69); J.A. Sanders, "Palestine Manuscripts 1947–1967," in: JBL, 86 (1967), 431–40 (includes a bibl. of the text-publications of Mss. discovered in Palestine from 1947 to Aug. 1, 1967). CONCORDANCES: K.G. Kuhn, Konkordanz zu den Qumrantexten (1960); idem, "Nachtraege zur 'Konkordanz zu den Qumrantexten,'" in: Revue de Qumran, 4 (1963), 163–234; A.M. Habermann, Měgilloṯ Miḍbar Yěhuda (1959), 3–175. WORKS: WAR SCROLL: J.P.M. van der Ploeg (ed. and tr.), Le Rouleau de la guerre (1959); Y. Yadin (ed.), Scroll of the War of the Sons of Light Against the Sons of Darkness, tr. by B. and C. Rabin (1962). THANKSGIVING PSALMS: M. Mansoor (ed. and tr.), Thanksgiving Hymns (1961); J. Licht, "Thanksgiving Scroll," in: Peruš Měgillat ha-Hodayot (1957); E. Kimron, "Language of the Psalms Scroll," in: Lèšonénu, 35 (1971/72), 99–116. PEšER ḤABAQQUQ: K. Elliger, Studien zum Habakuk-Kommentar vom Toten Meer (1953). MANUAL OF DISCIPLINE: P. Wernberg-Møller (ed. and tr.), Manual of Discipline (1957); J. Licht (ed.), Měgillat ha-Sěraḵim (1965; "Rule Scroll"); J. Maier (ed.), Texte vom Toten Meer… 2 vols. (1960), includes all the above scrolls. ZADOKITE DOCUMENTS: C. Rabin (ed. and tr.), Zadokite Documents (1954, 19582). STUDIES ON THE GRAMMAR IN THE DSS: E.Y. Kutscher, Ha-Lašon we-ha-Reqaʿ ha-Lěšoni šel Měgillat Yěšʿayahu ha-Šěléma (1959; "The Language and the Linguistic Background of the Isaiah Scroll"), includes summary in English; Ḥ. Yalon, Měgilloṯ Miḏbar Yěhuda; Divré Lašon… (1967; "Studies in the Dead Sea Scrolls, Philological Essay"), includes summary in English; M.Z. Qaddari, Ha-Ḥiyyuv bi-Lěšon ha-Měgillot ha-Gěnuzot (1968; "Semantic Fields in the Language of the DSS"), includes summary in English. SELECTED ARTICLES: B. Jongeling, "Les formes QTWL dans l'hébreu des manuscrits de Qumrân," in: Revue de Qumran, 1 (1958/59), 483–94; F.W. Bush, "Evidence from Milhamah and the Masoretic Text for a Penultimate Accent in Hebrew Verbal Forms," ibid., 2 (1959/60), 501–14 (against his conclusions see: E.Y. Kutscher above, 254–61); M.H. Goshen-Gottstein, "Philologische Miszellen zu den Qumrantexten," ibid., 2 (1959/60), 44–46; J.C. Greenfield, "The Root GBL in Mishnaic Hebrew and in the Hymnic Literature from Qumran," ibid., 2 (1959/60), 155–62; E.J. Revell, "The Order of the Elements in the Verbal Statement Clause in IQ Sereq," ibid., 3 (1961/62), 559–69; S.J. de Vries, "Syntax of Tenses and Interpretation in the Hodayoth," ibid., 5 (1964/66), 375–414; S. Lieberman, "The Discipline in the so-called Dead Sea Manual of Discipline," in: JBL, 71 (1952), 199–206; von N. Adler, "Die Bedeutung der Qumran Texte fuer die neutestamentliche Wissenschaft," in: Muenchener theologische Zeitschrift, 6 (1955), 286–301; J.P. de Menasce, "Iranien Naxcir," in: VT, 6 (1956), 213–4; W. Nauck, "Probleme des fruehchristlichen Amtsverstaendnisses," in: ZNW, 48 (1957), 200–20; Z. Ben-Ḥayyim, "Traditions in the Hebrew Language, with Special Reference to the Dead Sea Scrolls," in: Scripta Hierosolymitana, 4 (1958), 200–14; R. Meyer, "Spuren eines westsemitischen Praesens-Futur in den Texten von Chirbet Qumran," in: Gottes ist der Orient. Festschrift… O. Eissfeldt … (1959), 118–28; T. Leahy, "Studies in the Syntax of DSS," in: Biblica, 41 (1960), 135–57; A. Bendavid, Lěšon Miqra u-Lěšson Ḥaḵamim (19672), 80–94 (deals mainly with mutual influences of Aramaic, Greek, etc. and Qumran Hebrew). COPPER SCROLL: The language of the Copper Scroll, first published by J.M. Allegro (1960, 19642), which is close to Mishnaic Hebrew, was not dealt with in this article because the readings are not sure; see, however, B. Lurie, Měgillat ha-Něḥošet mi-Miḍbar Yěhuda (1963); H. Braun, Qumran und das Neue Testament, 2 vols. (1966); this important work systematically goes through the New Testament and compares it with the Qumran scrolls also in linguistic respects (lone translation); An index verborum of Qumran Hebrew only a Stellenregister, 368–83. See also bibliography at end of Dead Sea Scrolls article. ADD. BIBLIOGRAPHY: EDITIONS OF TEXTS: Y. Yadin, The Temple Scroll, Three Volumes and Supplement (1983); J.H. Charlesworth, ed., The Dead Sea Scrolls (1994– ); F. García Martínez and E.J.C. Tigchelaar (eds.), The Dead Sea Scrolls Study Edition (1997); E. Qimron, The Damascus Document Reconsidered (1992); idem, The Temple Scroll: A Critical Edition with Extensive Reconstructions (1996); idem and J. Strugnell, Miqsat Ma'ase Ha-Torah (DJD 10; 1994); idem & D.W. Parry, The Great Isaiah Scroll (1QIsaa): A New Edition (1999); J.K. Lefkovits, The Copper Scroll (3Q15): A Reevaluation (2000). GRAMMAR: E.Y. Kutscher, The Language and Linguistic Background of the Isaiah Scroll (IQIsaa) (1974), Eng. transl. of 1959 Heb. work; E. Qimron, The Hebrew of the Dead Sea Scrolls (1986); idem, "Observations on the History of Early Hebrew (1000 B.C.E.–200 C.E.) in the Light of the Dead Sea Documents," in: D. Diamant and U. Rappaport (eds.), The Dead Sea Scrolls: Forty Years of Research (1992), 349–62; S. Morag, "Qumran Hebrew: Some Typological Observations," in VT, 38 (1988), 148–64; M.S. Smith, The Origins and Development of the Waw-Consecutive (1991); T. Muraoka and J.F. Elwolde (eds.), The Hebrew of the Dead Sea Scrolls & Ben Sira (1997); idem, Sirach, Scrolls, & Sages (1999); idem, Diggers at the Well (2000); W.M. Schniedewind, "Qumran Hebrew as an Antilanguage," in: JBL, 118 (1999), 235–52; S. Weitzman, "Why Did the Qumran Community Write in Hebrew?" in: JAOS, 119 (1999), 35–45. CONCORDANCE: M.G. Abegg et al., The Dead Sea Scrolls Concordance: The Non-Biblical Texts (2002). MISHNAIC SPELLING: A. Bendavid, Lěšon Miqra u-Lěšon Ḥaḵamim, 1 (19672; to be used critically); E.Y. Kutscher, "Maẓẓav ha-Meḥqar šel Lešon Ḥazal," in: ʿEré ha-Millon he-Ḥadaš šel Sifrut Ḥazal (1971); idem, "Mi-Běʾayot ha-Millon he-Ḥadaš le-Sifrut Ḥazal," ibid,; (both articles contain many bibl. refs.); idem, "Mischnaisches Hebraeisch," in: Rocznik Orientalistyczny, 28 (1964), 35–48; idem, "Mittelhebraeisch und Juedisch-Aramaeisch imneuen Koehler-Baumgartner," in: B. Hartmann et al. (eds.), Hebraeische Wortforschung (1967), 158–75; idem, "Lěšonan šel ha-ʾIggerot ha-ʿIvriyyot wě-ha-ʾAramiyyot šel Bar Kosbaʾ u-Věné Doro," in: Lěšonénu, 26 (1962), 7–23; M.H. Segal, Mishnaic Hebrew and its Relation to Biblical Hebrew and to Aramaic (1909; repr. from JQR, 20 (1908), 647–737); Much material is to be found dispersed in the works of J.N. Epstein (below) especially in his Mavo lě-Sifrut ha-Tanna'im (1957); of S. Lieberman (below), especially in his Hellenism in Jewish Palestine (1950); and of H. Yalon (below). GRAMMAR: A. Bendavid (above); Z. Ben-Ḥayyim, "Traditions in the Hebrew Language with Special Reference to the Dead Sea Scrolls," in: Scripta Hierosolymitana, 4 (1958), 200–14; J.N. Epstein, Mavo lě-Sifrut ha-Tanna'im (1957), especially 1050, 1207–67; E.Y. Kutscher, "Biṣṣuaʿ Těnuʿot u, i be-Taʿtiqé ha-ʿIvrit ha-Miqraʾit ba-ʾAramit ha-Gelilit u-vi-Lěšon Ḥazal," in: E.Z. Melamed (ed.), Séfer Zikkaron lě-Binyamin De-Vries (1968), 218–51 (many printing errors corrected in rev. ed. (1971), idem, "Lěšon Ḥazal," in: S. Lieberman et al. (eds.), Séfer Hanoch Yalon (1963), 246–80; idem, "Meḥqarim bě-Diqduq Lěšon Ḥazal (lě-fi Kětav Yad Kaufmann)," in: Séfer Bar-Ilan, Qoveṣ he-ʿAsor 1956–1968, 2 (1968), 51–77; E. Porath, Lěšon Ḥaḵamim lě-fi Masorot Bavliyyot u-vě-ḵitvé Yad Yěšanim (1938); I. Yeivin, "Ha-Niqqud ha-Bavli u-Masoret ha-Lašon hamištaqqefet mimmenno" (1968; Typescript, unpublished Ph.D. dissertation Hebrew University, Jerusalem. Important); M.H. Segal, A Grammar of Mishnaic Hebrew (1927); M.Z. Segal, Diqduq Lěšon ha-Mišna (1936) includes bibliography (both outdated); Ḥ. Yalon, Mavo lě-Niqqud ha-Mišna (1964; very important work); idem, Pirqé Lašon (1971). Some material contained in this chapter is as yet unpublished. VOCABULARY: Ḥ. Albeck, Mavo la-Mišna (1959), 128–215; A. Bendavid (above). J.N. Epstein, Mavo lě-Sifrut ha-Tanna'im (1957); idem, Introduction to Amoraitic Literature (1962); idem, Mavo lě-Nusaḥ ha-Mišna (1964); (idem, many other works and articles published mainly in Tarbiẓ are very important); S. Lieberman, Ha-Yěrušalmi ki-Fěšuto (1934); idem, Greek in Jewish Palestine (1942); idem, Hellenism in Jewish Palestine (1950); idem, Tosefta (1970); idem, Tosefta ki-Fěšuta, 10 vols. (1955–67); idem, Tosefet Rišonim, 4 vols. (1937–39); idem, ʿAl ha-Yěrušalmi (1929); idem, "Roman Legal Institutions in early Rabbinics and in the Acta Martyrum," in: JQR, 35 (1944/45), 1–57; idem, "Palestine in the Third and Fourth Centuries," ibid., 36 (1945/46), 329–70; idem, many other articles (see T. Preschel, "Bibliografya šel Kitvé R. Šaul Lieberman," in: Hadoar, 42 (1963), 381–4); Ḥ. Yalon (above). DICTIONARIES: s.v. Aramaic; Ben Yehuda, Millon (contains the material of mishnaic Hebrew). Some material in this chapter is as yet unpublished. ADD. BIBLIOGRAPHY: Two collections of papers on MH are very useful and contain rich bibliography: M. Bar-Asher (ed.), Koveẓ Ma'amarim bi-Leshon Ḥazal, 1–2 (Jerusalem, 1972–80); M. Bar-Asher and S.E. Fassberg (eds.), Scripta Hierosolymitana 37: Studies in Mishnaic Hebrew (Jerusalem, 1988). A list of bibliography can also be found in M. Bar-Asher, "The Study of Mishnaic Hebrew Grammar – Achievements, Problems and Goals," in: Proceedings of the Ninth World Congress of Jewish Studies, Panel Sessions: Hebrew and Aramaic (Jerusalem, 1988), 30–37. GENERAL: M. Bar-Asher, "The Different Traditions of Mishnaic Hebrew," in: D.M. Golomb (ed.), "Working with No Data," Semitic and Egyptian Studies Presented to Thomas O. Lambdin (1987). MSS OF THE MISHNAH: M. Bar-Asher, The Traditions of Mishnaic Hebrew in the Communities of Italy: According to Ms. Paris 328–329, Eda Ve-Lashon, 6 (Jerusalem, 1980); Y. Bentolila, A French-Italian Tradition of Post-Biblical Hebrew, Eda Ve-Lashon, 14 (Jerusalem, 1989); G. Birnbaum, "Studies in the Phonology and Morphology of Mishnaic Hebrew According to Geniza Fragments," Ph.D. Thesis, Bar-Ilan University (1994); G. Haneman, The Morphology of Mishnaic Hebrew According to the Tradition of MS Parma (de Rossi 138) (1980); T. Zurawel, Maimonides' Tradition of Mishnaic Hebrew as Reflected in his Autograph Commentary to the Mishnah, Eda Ve-Lashon 25 (2004). OTHER TANNAITIC SOURCES: S. Naeh, "The Tannaic Hebrew in the Sifre according to Codex Vatican 66," Ph.D. Thesis, Hebrew University of Jerusalem (1989); H. Nathan, "The Linguistic Tradition of Codex Erfurt of the Tosefta," Ph.D. Thesis, Hebrew University of Jerusalem, Publication of the Faculty of Humanities, School of Advanced Studies (1984). AMORAIC HEBREW: Y. Breuer, The Hebrew in the Babylonian Talmud according to the Manuscripts of Tractate Pesahim (2002); E. Netanel, "Morphological Description of the Hebrew Verb in Jerusalem Talmud," Ph.D. Thesis, Hebrew University of Jerusalem (1995). OLD TRADITIONS: I. Eldar, The Hebrew Language Tradition in Medieval Ashkenaz (ca. 950–1350 C.E.), 1: Phonology and Vocalization; 2: Morphology, Eda ve-Lashon 4–5 (Jerusalem, 1978–79; Y. Kara, "Yemenite Traditions in Mishnaic Hebrew According to a 16th Century Manuscript," in: Leshonenu, 44:24–42; M. Ryzhik, "Italian Jewry's Traditions of Mishnaic Hebrew in the MSS of Mahzorim in the 14th–15th Centuries," Ph.D. Thesis, Hebrew University of Jerusalem (2001); O. Tirosh-Becker, "Rabbinic Hebrew Handed Down in Karaite Literature," Ph.D. Thesis, Hebrew University of Jerusalem (1999); I. Yeivin, The Hebrew Language Tradition as Reflected in the Babylonian Vocalization, 1–2 (1985). ORAL TRADITIONS: Y. Henshke, "The Hebrew Component in the Judeo-Arabic of Tunisia," Ph.D. Thesis, Hebrew University of Jerusalem (2000); K. Katz, The Hebrew Language Tradition of the Community of Djerba (Tunisia): The Phonology and the Morphology of the Verb, Eda ve-Lashon 2 (Jerusalem, 1977); idem, The Hebrew Language Tradition of the Aleppo Community: The Phonology, Eda ve-Lashon, 7 (Jerusalem, 1981); T. Kessar, Oral and Written Traditions of the Mishnah: Morphology of the Noun in the Yemenite Tradition, Eda ve-Lashon, 23 (Jerusalem, 2001); M. Mishor, "Ashkenazi Tradition – Toward a Method of Research," in: Massorot 3–4 (1989):87–128. SYNTAX: M. Azar, The Syntax of Mishnaic Hebrew (1995); N. Braverman, "Particles and Adverbs in Tannaitic Hebrew (Mishnah and Tosefta): A Syntactic Analysis," Ph.D. Thesis, Hebrew University of Jerusalem (1995); M. Mishor, "The Tense System in Tannaitic Hebrew," Ph.D. Thesis, Hebrew University of Jerusalem (1983). VOCABULARY: M. Moreshet, A Lexicon of the New Verbs in Tannaitic Hebrew (1980); Academy of the Hebrew Language, Ma'agarim (databases) CD-ROM (Jerusalem, 1998). ARAMAIC: I. Gluska, "The Influence of Aramaic on Mishnaic Hebrew," Ph.D. Thesis, Bar-Ilan University (1988). MEDIEVAL. THE LANGUAGE OF THE PIYYUT: Zunz, Lit. Poesie, 29–41; N. Chomsky, in: JQR, 75 (1967), 121; I. Davidson, in: Qoveṣ Maddaʿe ha-Yahadut (1926), 187–95; S. Lieberman, in: Sinai, 4 (1939), 221–50; M. Zulay in: YMḤSI, 6 (1945), 161–248; idem, in: Ḥ. Yalon (ed.), Qunṭěresim lě-ʿInyěné Lašon (1942–43), 1–4; idem, in: Moznayim, 16 (1943), 217–23; H. Yalon (ed.), Qunṭěresim lě-ʿInyěné Lašon (1942), 3–7, 51–55; S. Spiegel, in: Hadoar, 42 (1963), no. 23, 397–400; A. Mirsky, in: Ziḵronot ha-ʾAqademya la-Lašon ha-ʿIvrit (1956–57), 41–45; idem, in Lěšonénu, (1966), 296–304; Y. Kena'ani, Millon Qonqordanṣyoni li-Lešon ha-piyyuṭim (1936); K. Levias, in: Hadoar, 11 (1932), no. 33. SAADIAH GAON'S LANGUAGE: A. Ben Ezra, in: Séfer Alfenbein (1967), 33–43; idem, in: Horeb, 8 (1944), 135–7; 9 (1946), 176–85; 10 (1948), 295–318; M. Zulay, Ha-ʾAskola ha-Payyětanit šel Rav Sěʿadya Gaʾon (1964), 13–40; S. Abramson, in: Y.L. Eishman (ed.), Qoveṣ Rasag (1943), 677–88; idem, in; Sinai, 49 (1966), 133–245; C. Rabin, in: Saadya Studies (1943), 127–38. LANGUAGE OF THE HEBREW POETRY IN SPAIN: Schirmann, Sefarad, 2 (1956), 27–34; M. Medan, in: Lěšonénu, 17 (1951), 110–14; B. Klar, in: Meḥqarim wě-ʿIyyunim (1954), 174–9; S. Abramson, in: Ha-Kinnus ha-ʿOlami lě-Maddaʿé ha-Yahadut (1947–52), 274–8; A. Mirsky, in: Lěšonénu, 18 (1952–53), 97–103; S. Abramson, ibid., 11 (1941–43), 54–57; Y. Ratzaby, ibid., 21 (1957), 22–32; idem, in: ʾOṣar Yěhudé Sěfarad, 8 (1965), 11–16; in: Lěšonénu la-ʿAm, 2 (1969), no. 3–6; N. Allony, in: Lěšnénu, 11 (1941–43), 161–4; idem, in: ʾOṣar Yěhudé Sěfarad, 3 (1960), 15–48; idem, in: Sinai, 44 (1959), 152–69; 64 (1969), 12–35, 155–73; idem, in: Lěšonénu, 15 (1944), 161–72; D. Yarden, Diwan Šěmuʾel ha-Nagid (1966), 14–27; D. Yellin, in: JQR, 16 (1925/26), 272ff.; A. Mirsky, Širé Yiṣḥaq ḵalfon (1961), 40–44; D. Yellin, Ketavim Nivḥarim, 2 (1939), 319–30; idem, in: Tarbiz, 7 (1936), 314–19; Y. Ratzaby, in: Lěšonénu la-ʿAm (1959), 18. PROSE AND TRANSLATIONS WITH ARABIC INFLUENCE: A.S. Halkin, "The Medieval Jewish Attitude toward Hebrew," in: A. Altmann, Biblical and Other Studies (1963), 233–48; M. Gottstein, Taḥbirah u-Millonah šel ha-Lašon ha-ʿIvrit še-bi-Těḥum Hašpaʿatah šel ha-ʿArvit (1951); idem, in: Lešonenu 16 (1948–9), 156–163; B. Klar, in: Meḥqarim wě-ʿIyyunim (1954), 31–41; C. Rabin, in: Metsudah, 3–5 (1945), 158–70; J. Klatzkin, Oṣar ha-Munnaḥim ha-Pilosofiyyim 4 vols., (1928); I. Efros, in: JQR, 17 (1926/27), 129–64, 323–68; 20 (1929/30), 113–38; idem, Philosophical Terms in the Moreh Nebukim (1924); D.Z. Baneth, in: Tarbiz, 6 (1935), 10–40; G. Zarfati, Munněḥé ha-Matémaṭiqa ba-Sifrut ha-Maddaʿt ha-ʿIvrit šel Yěmé ha-Bénayim (1969); Z. Bacher (ed.), Séfer ha Šorašim (1896), 562–6; M. Wilensky (ed.), Sefer ha-Riqma, 2 (19642) 710–29; Z. Ben-Ḥayyim, in: Lěšonénu, 16 (1948–49), 156–63; M. Goshen, in: Tarbiz, 30 (1961), 385–95; D.Z. Baneth, ibid., 11 (1939/40), 260–70; 23 (1952), 111–76; N. Shapira, Lešonenu (26), 209; E.M. Lipschuetz, Ketabim 1 (1947), 203–209. ADD. BIBLIOGRAPHY: KARAITE HEBREW: A. Maman, "Ha-Ivrit shel Toviah b. Moses ha-Kara'i," M.A. Thesis, Hebrew University of Jerusalem) (1979); idem, "Karaites and Mishnaic Hebrew: Quotations and Usage," in: Leshonenu (1991), 221–68 (Heb.); idem, "Karaites and Mishnaic Hebrew: Quotations and Usage," in: M. Bar-Asher and S.E. Fassberg (eds.): Studies in Mishnaic Hebrew, Scripta Hyerosolomitana 37 (1998), 264–83; idem, "Karaite Hebrew, in: M. Polliack (ed.), Karaite Judaism, A Guide to Its History and Literary Sources (2003), 485–503. MAIMONIDES' LANGUAGE: B. Z, Bacher, in: ʿErḵé Midraš (1927), 324–36; P. Birnbaum, in: Hadoar, 21 (1942) 721f.; I.A. Zeidman, in: Sinai, 12 (1943), 428–38; 13 (1943), 96–101; M.Z. Qadari, Mi-Yrušat Lěšon Yěmé ha-Bénayim (1970). HEBREW IN FRANCE, GERMANY, AND POLAND: Y. Avineri, Ḥéḵal Raši, (1949); 3 (1956); 4 (1960); C. Rabin, in: Fourth World Congress of Jewish Studies, Papers, 2 (1968); S. Nobel, in: Lěšonénu, 23 (1959), 172–84, 216–69; S. Eidelberg, in: Lěšonénu la-ʿAm, 20 (1969), 120–7; Y. Avineri in: Metsudah, 2–4 (1945), 229–48. MODERN PERIOD Ziḵronot Waʿad ha-Lašon, 1–6 (1912–28); Lešonénu 1–35 (1913–71), (index volume, 1969); E.M. Lipschuetz, Vom lebendigen Hebraeisch (1920); idem, Kětavim, 2 (1949 50); Z. Har-Zahav, Lěšon Dorénu (1930), G. Bergstraesser, Einfuehrung in die semitischen Sprachen (1928), 47, 57ff.; S. Spiegel, Hebrew Reborn (1930); Y. Avineri, Millon Hiddušé Ḥ.N. Bialik (1936); idem, Héḵal Raši, 1–4 (1940–60); idem, Kibbušé ha-ʿIvrit bě-Yaménu (1946); Y. Epstein, Meḥqarim ba-Psiḵologya šel ha-Lašon wě-ha-Ḥinnuḵ ha-ʿIvri (1947); idem, Hegyoné Lašon (1947); Y. Klausner, Ha-Lašon ha-ʿIvrit Lašon Ḥayya (1949); R.W. Weiman, Native and Foreign Elements in a Language, a Study in Linguistics Applied to Modern Hebrew (1950); E. Rieger, Modern Hebrew (1953); Z. Ben-Ḥayyim, in: Lěšonénu la-ʿAm, (1953/54), 35–37; Ziḵronot ha-ʾAqademya la-Lašon ha-ʿIvrit, 1–17 (1954–71); A. Avrunin, Meḥqarim bi-Lěšon Bialik wě-Yalag (1954); Ḥ. Blanc in: Middle Eastern Affairs, vol. 5, pp. 385ff.; Ḥ. Rosén, Ha-ʿIvrit Šellanu (1956); R. Bachi, in: Scripta Hierosolymitana, 3 (1956); Y. Livni, Lašon ke-Hilḵatah (1957); W. Chomsky, HebrewThe Eternal Language (1957); idem, Ha-Lašon ha-ʿIvrit bě-Darḵé Hiṯpatteḥutah (1967); C. Rabin, in: Jewish Frontier (Sept. 1958), 11ff.; E.Y. Kutscher, Millim wě Tolědotéhen (1961); S. Morag, in: Lěšonénu la-ʿAm (1959/60), no. 95; C. Rabin, in: Jewish Frontier (June 1961); Z. Ben-Ḥayyim et al, in: Lěšonénu la-ʿAm (1960), no. 104; Ḥ.B. Rosen, A Textbook of Israeli Hebrew (1962); R. Sivan, Ṣurotu-Měgammot bě-Ḥiddušé ha-Lašon ha-ʿIvrit bi-Tequfat Teḥiyyatah (1964/65); Y. Avineri, Yad ha-Lašon (1964); R. Sappan, Darḵé ha-Sleng (1964); Z. Iggeret, Biʿur Koṣim mi-Kerem Lěšonénu (1964/65); R. Sappan, Millon ha-Sleng ha-Yisrěʾéli (1965); A. Bendavid, Lěšon Miqra u-Lěšon Haḵamim (I, 1967; II 1971); Ḥ. Rosén, ʿIvrit Ṭova (1967); M. Goshen-Gottstein, Mavo la-Millonaʾut šel ha-ʿIvrit ha-Hadaša (1969); M. Ben-Asher, Hitgabběšut ha-Diqduq ha-Normaṭivi (1969), S. Yeivin et al, in: Ariʾẹl, 25 (1969); Leqeṭ Teʿudot lě-Toledot Waʿad ha-Lašon weha-ʾAqademya la-Lašon ha-ʿIvrit56505730u-lě-Ḥiddušé ha-Dibbur ha-ʾIvri (1969/ 70); R. Sivan, in: Lěšonénu la-ʿAm (1970), 204–25; Munnaḥim ʿIvriyyim le-Miqṣoʿotehem (published by the Academy of the Hebrew Language, 1970). ADD. BIBLIOGRAPHY: Y. Bentolila, The Sociophonology of Hebrew as Spoken in a Rural Settlement of Moroccan Jews in the Negev (Hebrew University of Jerusalem, 1984)


Source: Encyclopaedia Judaica. © 2008 The Gale Group. All Rights Reserved.