A scholar's dictionary of Jewish Palestinian Aramaic.

• Author: Kaufman, Stephen A.
• Position: Michael Sokoloff

Lexicography is an art easily learned but seldom mastered. A comprehensive dictionary requires of its editor dozens of difficult decisions on every page. Moreover, different kinds of users have different needs, diverse enough that it often becomes impossible to satisfy them all in a single work. This is especially true in the case of rabbinic literature, wherein Hebrew and Aramaic materials of different dialects and periods are often interspersed in what can be a most chaotic fashion.

For several generations now, advanced students of the Aramaic and Hebrew dialects of rabbinic literature have been taught that the existing dictionaries of this literature are scientifically useless, inasmuch as they are based on inaccurate printed editions rather than on the best available manuscripts, and that words and meanings from the two different languages, or at least from the several distinct dialects of each, are inadequately distinguished within and across entries. Thus, however practical these works may be for students working with the traditional printed texts, they fail to present an accurate linguistic picture of any of the dialects in question, especially so after the publication of so many valuable treasures from the Cairo Genizah. The impressive, massive Dictionary of Jewish Palestinian Aramaic by Michael Sokoloff, Professor of Hebrew and Semitic Languages at Bar-Ilan University, constitutes an attempt to begin to remedy that situation on the Aramaic side. It is to be followed by a companion volume on Jewish Babylonian Aramaic (JBA), which Sokoloff is currently working on under the auspices of the Comprehensive Aramaic Lexicon Project.(1) This dictionary covers the textual material from several closely related but hardly identical dialects of Jewish Palestinian Aramaic (JPA), in ten major divisions:

1. Inscriptions, mostly from the remains of

   synagogues.

2. Targums; specifically: Targum Neofiti, the

   Genizah fragments of the Palestinian targums,

   and the Fragment Targums. III. Palestinian Midrash.

4. Palestinian Talmud.

5. Fragments of halachic literature of the Gaonic

   period.

6. Liturgical poetry (piyyutim). (An addendum

   includes a selection of material from five new

   poems discovered after the main text was set

   for press. VII. Fragmentary papyri with the remains of letters

   and documents from 5th cent. C.E. Egypt. VIII. Amulets--both originals from Palestine and

   manuscripts from the Genizah.

9. Ketubbot from the Genizah--Arabic period

   documents preserving old Palestinian terminology.

10. Masoretic notes on medieval Tiberian Bible

    codices.

    From the point of view of dialect, these materials can be divided roughly into three major divisions: inscriptional (I), targumic (II), and "Galilean" (III-X). Proper nouns are not generally included, though some lexically interesting ones are (e.g., [UNKNOWN TEXT OMITTED]`Leviathan' is given, but not the ubiquitous, [UNKNOWN TEXT OMITTED] Pharaoh').

    Clearly, most of this material was not available to lexicographers prior to the Cairo Genizah discoveries and publications, reason enough for a new lexicon; but to create such a lexicon required a totally new approach. One could not simply extract a list of headwords and citations from the older dictionaries and modify them slightly according to a shoe box full of additional citation slips, the traditional procedure followed by most dictionaries in most fields, be they monolingual or bilingual. Instead, Sokoloff first had the entire text corpus input into the Bar-Ilan computer. Then, using specially modified programs originally developed for the Bar-Ilan responsa project, the texts were lexically tagged by the computer (by no means a trivial job for a consonantally written Semitic language), and key-word-in-context (KWIC) concordances were generated, corrected, and generated again. Finally, the dictionary entries themselves were written using the data assembled in the KWIC concordances. The entire process took about ten years--a very short time, as such projects go.

    The result is a lexicon of the classical kind, wherein citations of varying length and accompanying English translations justify each meaning. Cognate forms are regularly adduced from the other late Palestinian Aramaic dialects (Samaritan and Christian Patestinian Aramaic) and Syriac, and notes to discussions of words in the scholarly literature are impressively extensive. Semitic dictionaries have traditionally been arranged by triliteral root. On the other hand, the modem Akkadian lexica (CAD, AHw) as well as dictionaries of Modem Hebrew and even some classical Arabic dictionaries,(2) are ordered purely alphabetically. Each sequence has its own set of advantages and disadvantages, and little is to be gained by criticizing one choice or the other.(3) Whatever system is chosen, one must allow for homographs in some rational manner. DJPA uses an alphabetic arrangement; homographs are ordered by part of speech, and are numerically distinguished only when both the part of speech and consonantal orthography are identical. Nouns are listed in the absolute state--an appropriate choice for a Western Aramaic dialect. But what is the spelling of each headword to be? Jewish texts are particularly problematic for both lexicographer and reader because of rampant inconsistencies in the use of vowel letters. The lexicographer is thus required to choose from among several possible spellings for many if not most headwords. Sokoloff's choice of spelling practice is consistent with the philosophy underlying the work as a whole: "the entry header itself is spelled in accordance with what has been shown to have been the original JPA practice" (p. 6). Cross-references under other common spellings are used, but not regularly. Thus, by the very choice of headword spelling, Sokoloff has demonstrated his implicit desire to produce not a reference work for students but, rather, first and foremost, an accurate presentation of the lexicon and orthography of his dialects. The choice and extent of selected citations and translations, the absence of references for clear scribal errors, and the limited adduction of support from cognate dialects all comport with this lexicographic philosophy. Such a presentation constitutes a legitimate scholarly enterprise. It should not be criticized for failing to be what it does not try to be. It will not easily serve the needs of a beginning student.

    On the other hand, one is entitled to question whether DJPA lives up to its own standards and does present an accurate picture of the attested lexical stock of Jewish Palestinian Aramaic of the Byzantine Period. Here we must express some mild dissatisfaction, in particular as regards targumic texts:

    1. Although it is nowhere mentioned in the very limited

       introduction to the work, the fact is that the

       extensive marginalia to Targum Neofiti to the

       Pentateuch (TNGI) were not fully incorporated

       into the database from which the dictionary was

       created--only citations deemed to be lexically interesting

       were included. Our notes below attempt

       to remedy this lacuna.(4) b) The Paris 110 manuscript of the Fragment Targum

       is completely included, but that text has now

       been shown to be heavily influenced by European

       traditions,(5) so forms unique to it must not be ascribed

       to this dialect. On the other hand, Sokoloff

       is quite correct in having excluded the Pseudo-Jonathan

       Targum from consideration. Although it

       incorporates much Palestinian targum material, it

       is at core a late composite work in a totally different

       dialect.(6)

       Equally, one must ask if the chosen system of headword spelling, ordering, and cross-referencing provides a consistent and maximal guide to the entirety of the lexicon. Generally the situation is quite satisfactory in this regard, but there are some disturbing inconsistencies. Unfortunately, the introduction is lamentably brief, as a result of which the lexicographical principles governing the dictionary, including those that might account for the inconsistencies, are never explicitly stated. In all such cases, it is quite possible that a logical and correct explanation for the apparently inconsistent approach may exist, but if so the user searches in vain for it. Examples: a) Words that are conceptually a part of other headwords are treated in four different ways: 1) They can have a separate entry (e.g., [UNKNOWN TEXT OMITTED] adj., `exalted', actually a derived stem passive participle of the verb [UNKNOWN TEXT OMITTED!; or the hapax [UNKNOWN TEXT OMITTED] `dispersed one', a simple passive participle translating a Hebrew nifal participle. 2) They can be listed simply as a part of the main entry (e.g., [UNKNOWN TEXT OMITTED! `destroyed' is not listed, only given as passive participle under the verb; [UNKNOWN TEXT OMITTED] why is this different from the previous case?). 3) There is a very strange amalgam of (1) and (2) where we find an abbreviated main entry, with a cross reference to the base word (e.g., [UNKNOWN TEXT OMITTED] adj. `pregnant' [p. 321], which is, of course, a passive participle as well)! 4) The same word can appear twice, as both (1) and (2): Targumic [UNKNOWN TEXT OMITTED] (Lev. 21:7) is listed as a pael p.p. under [UNKNOWN TEXT OMITTED] vb., but Galilean attestations are treated under the separate headword, [UNKNOWN TEXT OMITTED] `divorcee'.(7)

    2. The placement and treatment of compounds always present problems for the lexicographer, but the user expects consistency, whatever approach is chosen. Again in this case, however, inconsistencies are never accounted for. It is by no means clear, for example, why some compound prepositions are listed as independent words, others cited as indented words after the main entry of their first component, and yet others cited only as sub-meanings of the second component. An example of the last is, [UNKNOWN TEXT OMITTED] `in view of' (booked under), [UNKNOWN TEXT OMITTED] as opposed to, say, [UNKNOWN TEXT OMITTED] towards', which is a separate entry...