Reassessing the Impact of Barthélemy’s Devanciers, 40 years later

3. Patterns of Translation Technique and Efforts at

“Revision”

B demonstrated that the Naal ever Minor Prophets translation exhibited a fairly consistent technique with respect to various linguistic features (see above, n. 11), and he attempted to trace these and related features in other early translation literature. In section F of his edition of the Minor Prophets material, Tov presents extensive detail on “Translation Technique, Orthographic Peculiarities and Textual Relations” (99–158), and elsewhere comments that “as a rule, [the translation] is rather consistent, so that its vocabulary and system of translation can be identified in other [fragmentary] instances as well” (83). Other authors have proposed additional linguistic criteria in studies of “καίγε-Theodotion” in other sections of Greek Jewish scriptures.

This awareness of patterns not only served to recreate the general parameters of καίγε-Theodotionic techniques, whether centered in an individual or a “school,” but also provided more systematic data for exploring the relationship of the old καίγε version(s) to manuscripts and variations within the LXX/OG traditions and to the materials collected by Origen in his massive Hexapla. For example, as has already been noted, Barthélemy and Tov argue that the evidence suggests that in the Minor Prophets the newly recovered καίγε translation is a revision of an older Greek text that can be reconstructed from the extant LXX/OG manuscripts and witnesses. They also argue that the Hebrew Vorlage behind the καίγε revisions is somewhat closer to what became the Hebrew MT than to the LXX/OG (lost) Hebrew Vorlage (Vorlagen?!). The arguments and impressions on which such reconstructions rest are quite subtle, and in the world of the first century b.c.e. that seems to have known variations in Hebrew copies (as attested in the DSS), not entirely persuasive. The activity of translation, especially where some level of “literalism” is desired, presents limited possibilities for any word or construction. How one establishes priority in such circumstances is problematic, and often rests as much on assumptions (what Vorlage was older, whether woodenness is more likely to give way to idiomatic or vice versa) as on clear features in the available evidence. Similarly, any ancient impetus to keep the translation close to the current supposed “parent text” (as with Aquila) or to provide a more idiomatic and/or varied flavor (as with Symmachus) will depend on a variety of factors (e.g. attitude to the texts, availability of variant forms), most of them lost to us. Barthélemy’s work has opened these doors more widely to contemporary scholarship, even when he did not recognize all of the implications.

This increased awareness of patterned translation—of more or less predictable translation technique—can only benefit the study of the preserved materials (manuscripts and other witnesses), especially in complex situations such as presented by the Greek witnesses to Samuel-Kings. None of this is particularly new, in principle, but the ability to present in more detail the various features of καίγε-Theodotion serves as a catalyst to more precise analysis of the data. Several relevant studies have been produced, many of them by students of Frank M. Cross, including at least one that B himself reviewed quite favorably in print.³¹ This focus has spilled over into studies of deuterocanonical and parabiblical materials as well, such as  Sirach and Tobit, and deserves to be explored more in such texts as the Testaments of the Twelve Patriarchs.³²

4. Samuel-Kings as a Test Case, Translation Technique

as a Criterion, and Possible Hexaplaric Confusions

Probably the most fruitful area in which B’s research has developed and is still developing is the study of the Greek witnesses to Samuel-Kings. This, indeed, was the subject of the aforementioned 1972 IOSCS/SBL Symposium (above, n. 4) to which B himself was invited, although as it turned out, he was unable to attend. As noted above, four main papers were pre-published for that occasion: Tov, Barthélemy, Muraoka, Cross. Unfortunately, an official record of the discussions has not, to my knowledge, been preserved although B’s prepared response published subsequently in Études is of some help, since he addresses several of the issues raised.³³

A significant part of the argument in Devanciers is devoted to close analysis of portions of Samuel-Kings. The basic issues were already well known to students of Greek Jewish scriptures, and had been laid out most clearly by H. St. J. Thackeray in his 1920 Schweich Lectures, The Septuagint and Jewish Worship (1921). Discussion is complicated by the fact that the pertinent sections of Samuel-Kings (Greek 1-4 Reigns) in the surviving Greek manuscripts do not divide neatly at obvious points, and thus have received from Thackeray what seem, at first glance, to be rather cryptic designations (using Greek letters/numbers for each book of Reigns, alpha to gamma) to indicate obvious changes in translational styles and textual affinities. For those unacquainted with this coding, the use of English designations might be worth attempting:

“1S” (α) is unambiguous, for the entirety of 1 Samuel (Reigns α)

“2S1” (ββ) represents 2 Sam 1:1–11:1 (Reigns β, part one);

“2S+” (βγ) covers 2 Sam 11:2 through 1 Kgs 2:11 (Reigns β, part two, through Reigns γ, part one);

“1K2” (γγ) is 1 Kgs 2:12–21:43 (Reigns γ, part two, or most of the rest of that book); and

“2K+” (γδ) is the remainder, from 1 Kings 22 through 2 Kings (Reigns γ, part three, through Reigns δ).

B’s main focus was on our 2S+ (Thackeray’s βγ), where he found the καίγε translation features in the majority of the OG manuscripts, although not in the variant text represented by manuscripts boc₂e₂ (using the sigla of the Larger Cambridge Septuagint). A similar situation was evidenced in 2K+ (γδ). But in the other sections of Samuel-Kings, the extant Greek manuscripts did not exhibit the καίγε characteristics. Whereas Thackeray had conjectured that this phenomenon of inconsistency in the main body of Greek manuscripts had been caused by two different sets of translators working on the different sections at different times and for different reasons (an abridged translation containing 1S, 2S1, and 1K2 from Alexandria, later patched with 2S+ and 2K+ from an Ephesian Jewish translator), B argued that Origen’s Hexapla had inadvertently created the situation by mixing manuscripts of the καίγε recension with manuscripts of the older OG to produce what came to dominate the later copies. Origen lived at a time when scrolls and small codices were the norm, and thus such a confused situation can easily be imagined for a collector of available texts. Fortunately, portions of the minority OG text also survived elsewhere in the Hexapla for 2S+ and 2K+ (and in manuscripts boc₂e₂), although the καίγε version has been lost for the remainder of Samuel-Kings.³⁴

Previous research on this material had identified manuscripts boc₂e₂ as “Lucianic,” representing a supposed early fourth century revision of the Greek scriptures attributed to the Antiochian martyr Lucian (died 311). B’s understanding of the situation led him to see this supposed “Lucianic” material as the only manuscript remnants of the OG for the sections 2S+ and 2K+. In this connection he denied the accuracy of the “Lucianic” label in Samuel-Kings, and called into question the existence of the “Lucianic recension” in general. This stirred up a hornet's nest of response, as can be seen from Tov’s “State of the Question” essay in the 1972 Symposium volume.³⁵ B’s detailed contribution to the Symposium nuanced his position while in general reaffirming its main points—while the situation is certainly more complex than presented in Devanciers, and boc₂e₂ do show themselves to represent a revised form of the OG in the relevant sections (and elsewhere as well; see e.g. Bernard Taylor’s studies, n. 35 above), B explains at some length how “Lucian’s recension” was an invention of much later Christian wishful thinking.³⁶

In his second thoughts about the existence of a “Lucianic recension,” B once more illustrated an area of his research that deserves more attention and more emulation, namely, historical contextualization of textual develop­ments. After exploring the history of (vague) references to “Lucian” as someone who did something noteworthy with biblical texts, B argued that there is no solid reason to ascribe a “recension,” or even an “edition” of the Christian scriptures to Lucian, and lots of reasons to think he was honored with that distinction by later admirers to enhance his image. Whether B is correct in this judgment, and whether it really makes much difference beyond the question of labels, B once more opened a door to further exploration. He had already done this with his attempt to identify the work of the καίγε-Theodotion school with first century c.e. Palestinian proto-rabbinic efforts—probably too optimistically. In the related area of the quotations in the Greek manuscripts of Philo, where some witnesses have the expected LXX/OG text form while a few exhibit an Aquila-like text, B has provided evidence that the insertion of the latter may have been the work of a Jewish reviser, operating in the area of Caesarea around the time of Origen (see above, n. 23). B also attempted to understand Origen’s procedures in compiling the Hexapla, as part of B's solution to the Samuel-Kings textual confusion.³⁷ Of course, other explanations for all these situations are also possible, but the mere fact of attempting to find suitable historical contexts for such developments is an advance over being unaware of the possibilities.

The larger issue here is, in part, the attempt to trace the history and influence of the versions of scriptural works that can be identified by paying attention to translation technique, especially καίγε-Theodotion and Aquila. B envisioned a closely related collection of καίγε-Theodotion revisions of LXX/OG emerging from a Palestinian “school of translators” in the first century c.e.. More likely, at least some of these efforts (the Minor Prophets) took place at least two generations earlier, in the first century b.c.e., in a period in which textual fluidity among Hebrew texts is now well documented, and in which the technology of scroll production at that time makes it difficult to imagine that a consistent text of major portions of what may have been considered “holy scripture” could easily be transmitted and maintained. Next B gave us a glimpse of these καίγε-Theodotion materials providing a base for the activities of “Aquila” (perhaps also a translational “school”?) in the second quarter of the second century, and around the same time being used by Justin, if the manuscripts of Justin can be trusted (see above, n. 23). Meanwhile, Aquila’s popularity led to the work of B’s Caesarean rabbinic emender of Philo’s texts, and both Aquila and some copies of καίγε-Theodotion texts fell into Origen’s willing hands for inclusion in the Hexapla, with some confusion along the way—not unusual or unexpected in that transitional period from scroll to codex. Aquila survives in Jewish circles beyond the time of Justinian into the sixth century, while the fate of the καίγε-Theodotion translations is less clear.³⁸ Indeed, to some extent the work of Origen must have been a major event that contributed both to the memory of these versions and to their complex survival histories and/or disappearance.