Review of Lust et al., A Greek-English Lexicon of the Septuagint, Pts. 1-2

J. Lust, E. Eynikel, and K. Hauspie, with the collaboration of G. Chamberlain. A Greek-English Lexicon of the Septuagint. Pt. 1. Stuttgart: Deutsche Bibelgesellschaft, 1992. ISBN: 3-438-05125-7. Pp. LIII+217. US $26.25.

J. Lust, E. Eynikel, and K. Hauspie. A Greek-English Lexicon of the Septuagint. Pt. 2. Stuttgart: Deutsche Bibelgesellschaft, 1996. ISBN: 3-438-05126-7. Pp. LXVI+311. US $33.75.

The Need for a Lexicon of the LXX

1. LXX scholars have long dreamed of having a LXX lexicon. From its inception, the long-term goal behind the creation of the CATSS project based at the University of Pennsylvania and jointly led by Robert Kraft of the University of Pennsylvania and Emanuel Tov of Hebrew University was the preparation of an electronic lexicon (Kraft and Tov, eds. 1986: 2).

2. In the interim, scholars have had to make do as best they could. The first English lexicon sensitive to LXX nuances was that of Abbott-Smith (Abbott-Smith 1937), but coverage did not extend beyond NT vocabulary, leaving some 60% (his estimate) of LXX words not included. At first blush the abridgment of the standard classical Greek lexicon (the so-called "Intermediate" Lexicon) prepared by Liddell and Scott seems to be a viable alternative (Liddell and Scott 1871). However, the "Advertisement" bound following the title page says, "The Abridgment . . . is intended chiefly for use in Schools." While care was taken to include all NT words, clearly the LXX lay outside the purview of the volume, since no abbreviation is even listed. So the volume lacks the distinctive LXX vocabulary.

3. Thus the only option has been the full LSJ (Liddell and Scott 1925-1940). There are several problems with this, not the least of which is having to look through so much material to find any given word. Beyond that, the focus is naturally on classical/Attic forms, and in that context it is not surprising that LXX/koine forms fare poorly, as LXX scholars have noted as each subsequent edition of the volume has been reviewed. Not the least problem in this regard is the subsuming of koine forms under a rubric such as "later." For example, <grc>r(h/ssw</grc> is listed under <grc>r(h/gnumi</grc>. To the extent that koine is treated as a step-child of classical forms, to that extent it is prevented from standing in its own right as a legitimate, albeit later, form of the language.

4. I do not mean to single out LSJ in this regard. When classical/Attic Greek was taught in high schools and colleges, the normal route to both NT and LXX Greek was through classical studies, which were ipso facto normative, and this is still reflected in standard reference works: for headwords in their concordance Hatch and Redpath usually list the Attic word first, followed by the koine form(s) (where different) (Hatch and Redpath 1998); Conybeare and Stock wrote their introductory grammar to bridge from classical to koine, which explains why it appears so unhelpful for those who know NT but not classical Greek (Conybeare and Stock 1905); and Thackeray in his grammar is similarly focused (Thackeray 1909).

5. The one work to which those who had access could turn was Schleusner's Thesaurus (Schleusner 1829). Since many LXX scholars may never have seen this work, it is worthwhile to point out its drawbacks. First, it was in Latin. Second, Schleusner was far more interested in the underlying Hebrew than in LXX Greek per se, to the extent that it was essentially a study of Hebrew-Greek parallels. Even if it were in print, it would do little to assuage the desire for a LXX lexicon, though to some extent it anticipated current interest in translation technique. In a real sense, then, the lexicon under review is sui generis and the first English lexicon dedicated to the LXX.

The Lexicon and the CATSS Database

6. As is inevitable, there is a prehistory that needs to be borne in mind. First, the lexicon is based on the CATSS database (cf. Kraft and Tov, eds. 1986: 1-12 for a description of the project). Early on in the CATSS project a morphological analysis of the Rahlfs text was prepared (CATSS Morph LXX). While the project has proceeded a little beyond that point by compiling variant readings for some of the books, it has not reached the original goal of creating an electronic lexicon. Thus it is appropriate that it should spawn a lexicon, even if in hardcopy.

7. Unfortunately, although (a preliminary release of) the morphological analysis is available both on CD and the Internet, there are still significant problems yet to be corrected. Many of these corrections were found in the preparation of my Analytical Lexicon (Taylor 1994), but they have yet to be fully integrated into the database. Though the compilers of the work under review also corrected much, in important ways the lexicon is heir to these errors, not the least effect of which is to cast doubt on the word counts, especially for those that occur frequently. One way to deal with this problem would be to check the entries against Hatch and Redpath, though this work is not without error, either. This is not to deny the usefulness of the numbers, only their ultimate accuracy, and there is no way to predict in advance just which entries might be affected.

The Textual Base of the Lexicon

8. Behind the CATSS database lies the TLG machine-readable Rahlfs text, and it is upon this text that the lexicon is based. The Rahlfs text is the only complete (semi-)critical edition of the Old Greek text. The Swete text (Swete, ed., 1887-1912), the basis for the Larger Cambridge Septuagint (Brooke et al., eds. 1906-), was a diplomatic edition based on MS B (Codex Vaticanus). As the compilers of the lexicon recognize, first choice would have been the Göttingen text (Ziegler et al., eds. 1931-) had it been complete, since it is a full critical edition.

9. As is common practice, proper nouns, except where they are transliterations of Hebrew common nouns, are not included. While this decision is eminently defensible, it does leave a fruitful area yet to be adequately studied. Tracing proper names through manuscripts is a key indicator of family groupings, and some indication of etymology would have been a useful addition.

Description and Critique of the Lexicon

10. Individual entries are listed alphabetically, not grouped together in semantic domains.

11. Each word has its morphological tag. These are single letter tags following the CATSS system. I for one would prefer at least two-letter tags, since they are much more intuitive. The problem arises over duplication of initial letters for some of the parts of speech such as "adjective" and "adverb" ("A" and "D" respectively), and "preposition," "pronoun" and "particle" ("P," "R," and "X"). Alternatives for the former could be "Aj" and "Av"; for the latter "Pe," "Po," and "Pa"--or even "Pre," "Pro," and "Par"--would be preferable.

12. Word frequencies are tabulated in six categories: Pentateuch, Early Prophets plus Chronicles, Later Prophets, the Hagiographa (excluding Chronicles), the books not found in the modern Tanakh, and the total of the preceding five. This way the protocanonical books are for the most part kept separate from the deuterocanonical ones. The exceptions are the additions to books such as Esther and Daniel, which are not differentiated in the statistics from those sections present in Hebrew. Beyond that, no distinction is made between translations and original Greek compositions in the Apocrypha.

13. In the interest of bringing the work to completion as quickly as possible, the compilers opted for glosses or translation equivalents rather than descriptions of meaning. Thus, <grc>a)rh/n</grc> is glossed as "lamb" (in the sense of "a lamb," not the meat), whereas a description would be something like "the young offspring of (a) sheep."

14. The English that is used is British or International rather than American, as can be seen from the spelling. This fact must be kept in mind when reading word definitions. For instance, <grc>a)nabra/ssw</grc> is listed as "to throw up, to reject [<grc>tina</grc>]," citing Wis 10:19. It must be understood that this is not colloquial American English.

15. Up to five references are listed in order of appearance. If the word occurs relatively frequently the list may not extend beyond the Pentateuch, and for common words such as <grc>e)/rxomai</grc>, it may not go beyond Genesis. In hindsight, it might have been useful to cite references from various sections of the LXX rather than concentrate them in one area. For each significantly different translation at least one additional reference is supplied.

16. At first glance it appears as though prepositions are undifferentiated as to meaning in relation to their use with different cases. In fact, however, case is indicated by the form of <grc>tis</grc> (such as <grc>tinos</grc>) as the headword for the relevant portion of the entry.

17. Where appropriate, words are classified in terms of four further categories, although which falls into what category is not specifically marked except as the wording itself indicates (such as "neol."; see below, par. 23).

18. The first relates to those words which can with a degree of certainty be labeled as classical Greek, as well as forms that require special translation. Just what this latter category may include is not elucidated. For example:

<grc>neotto/s</grc>, -<grc>ou=</grc>
4 Mc 14,15
Att. form of <grc>neosso/s</grc>; young bird, nestling

19. The second category includes those expressions that are not felicitous in Greek but rather literal renderings of Hebrew (and presumably Aramaic) idioms. Apparently this includes Semitisms. For instance:

<grc>kinna/mwmon</grc>, -<grc>ou</grc> . . .
Semit. loanword (Hebr. <heb>qnmwn</heb>) . . .

20. Category three contains those passages where the Greek text itself may be corrupt. In the first instance this relates to the underlying "LXX/OG" text, not the Rahlfs text. In practice there are few of these, because of the leveling effect of the Rahlfs text, though that text itself is not without its typographical errors:

<grc>a)potruga/w</grc> . . .
. . . *Am 6,1 <grc>a)petru/ghsan</grc> they have gathered or plucked (metaph.) corr.? . . .

21. The last category is for those places where the LXX/Rahlfs text differs from the MT. This may be due to misreading, or reading the same text differently, or a divergent text. Thus:

<grc>e)xqro/s</grc> . . .
. . . *Jb 22,25 <grc>a)po\ e)xqrw=n</grc> <heb>m</heb>/<heb>cry</heb>- from enemies for MT <heb>bcry</heb> gold . . .

22. In the second and third categories the beginning of the discussion is marked by an asterisk, and the particular reference relative to the discussion is indicated. Such discussions are not intended to be exhaustive of all possible examples.

23. When a word is deemed to have originated in the world of the LXX, it is marked as a neologism ["neol."]. When it is also found in contemporary literature from the time of Polybius on, a question mark is added to the neologism label ["neol.?"]. The intended implication is that the word may not have been used before the time of the LXX, and may well have been coined in the translation process. At best this is a value judgment in the face of not a few variables, and Lust is sensitive to the implications. However, despite the uncertainties in the dating of both the LXX and the various papyri, one is able to make some useful comparisons within the established parameters, and these are included in the lexicon.

24. An example of "neol." is <grc>e)cole/qreuma</grc>, and an example of "neol.?" is <grc>e)coleqreu/w</grc>. In the entry for the former word, LSJ lists only LXX 1 Ki 15:21 (= 1 Sam). For the latter word, LSJ cites the LXX, the NT, and one reference to a variant reading containing a form of this word in Josephus, Ant. 8.11.1. Thus all of the references are LXX or post-LXX .

25. In the case of compound verbs, Lust's Introduction is misleading. It states: "compound verbs are referred to under the simple form." What is meant is that any simple form of a verb with one or more compounds found in the LXX is followed by the list of prefixes for the associated words. Each compound verb has its own entry apart from the root. If the simple form represented by a compound is not found in the LXX, a dummy entry is provided without statistics, etc., and prefixes of related compounds are listed. Note, however, that compound forms are not indicated in any way such as hyphenation. If unsure whether a word is a compound or not, the only avenue is to check for a listing of the putative simple form. <grc>e)/xw</grc> is an example of a verb that occurs in the simple form, and so has all of the usual entries, but also takes compounds, in this case 18 different prefixes or combinations.

26. The three compounds <grc>ei)sodia/zw</grc>, <grc>e)codia/zw</grc>, and <grc>e)podia/zw</grc> all share the root -<grc>odia/zw</grc>, which does not occur in the LXX, so a dummy entry is provided that simply lists the three prefixes. Note in this connection that the prepositions are listed in the form they take when attached, e.g., in the above case <grc>e)c</grc>- not <grc>e)k</grc>-.

27. Entries conclude with bibliographical references (where they exist) to articles specifically related to the particular word under consideration.

28. The lexicon is complete in the two volumes. Originally it was planned to include a third volume of morphology, but this was rendered unnecessary by the reviewer's Analytical Lexicon (Taylor 1994). Instead, they plan to concentrate on a third volume dealing with non-LXX variants such as are found in the apparatuses of various critical editions (vol. 2, p. I).

29. The first volume shows all the hallmarks of having been prepared in a hurry. It comes with a small corrigenda, but this does not begin to address the problems. The Introduction has more than its fair share of typographical errors and misspellings. While this has not carried over into the text itself with the same intensity, it also has errors, including the omission of some Greek words and misparsing of others. However, the compilers recognize this and plan to bring out a second edition of volume one in the near future. In the meantime, volume two shows the hallmarks of more careful work. In addition, the bibliography of the first volume is carried to completion. An example of an omission is <grc>ei)=mi</grc> / <grc>i)e/nai</grc> (=ibo), even though compound forms do occur. An example of misparsing, carried over from CATSS, is the parsing of <grc>e)sxatogh/rws</grc> as an adverb instead of a second declension Attic adjective.

30. At present the volumes are available only as separate paperback editions. The binding permits the books to lie flat, and they are a pleasure to use. For extended use in an environment such as a library, though, the volumes in the present form would not fare well. A single hardcover edition would surely be popular in extended wear situations.

The Approach of the Lexicon

31. So much for the preparation of the volume. What of the underlying philosophy? Lust addresses the major issues in Part II of the Introduction to volume one under the rubric "Translation Greek." In koine Greek the LXX is for the most part a translation and not an original composition, hence Lust's choice of the title "Translation Greek." Anyone who attempts to translate the LXX--and compiling a lexicon involves translating--is pulled in two essentially opposite directions: towards the Hebrew, since it is a translation, and towards the Greek qua Greek, since the finished product is a Greek document with a history in many circles independent of the Hebrew. Naturally opinion is divided over the prudent approach to the topic. While sympathetic to the latter approach, Lust chose to concentrate the team's work in the area of translation. (Ironically, Muraoka, who is also in the process of creating a LXX lexicon [Muraoka 1993], chose the opposite as his emphasis.)

Conclusion

32. Problems aside, the preparation of the volumes represents a major stride forward in LXX studies. I look back on my first attempts to read the LXX using the available New Testament lexica, even with the addition of the Intermediate LSJ, and remember the heightened frustration I felt when that did little to help, though it appeared to promise much. Now I luxuriate in a lexicon devoted solely to the LXX. Lexicon compilers in the future will do differently, and hopefully even better by building on this foundation, but LEH did it first and did it well within a relatively short period of time.

© TC: A Journal of Biblical Textual Criticism, 1998.