Jewish Texts and History • The Septuagint

The term Septuagint is derived from the Latin septuaginta, an abbreviation of interpretatio secundum septuaginta seniores, that is, ‘the interpretation of the seventy elders’ ...

Contemporary scholarship uses the term Septuagint to denote four notionally distinct corpora: (1) a Greek translation of the five books of Moses (the Pentateuch) believed to have been undertaken in third-century bce Alexandria; (2) the so-called Alexandrian Bible, a Jewish scriptural corpus in Greek dating to the late Hellenistic period; (3) the Greek Old Testament, a Christian corpus comprised of the books of the Hebrew Bible as well as the so-called Deuterocanonical books (or Apocrypha); and (4) the earliest recoverable translations of the Jewish Scriptures into Greek, often referred to as the Old Greek versions. While there is substantive overlap between these four corpora, each raises its own literary and historical issues, and each has a distinct cultural value ...

The Greek phrase κατὰ τοὺς ἑβδομήκοντα, ‘according to the Seventy’, began appearing in Christian biblical manuscripts around the fourth century ce (Dines 2004: 1). It alludes to a Jewish legend according to which a Greek version of the Law was commissioned by Ptolemy II Philadelphus (r. 283–246 bce), and undertaken by seventy (or seventy-two) Judean scholars. If the testimony attributed to the Jewish philosopher Aristobulus (c.170 bce) is genuine (ap. Eusebius, Praep. Evang. 13.12.2), then the main features of the legend were already taking shape within a century of the supposed events.

The Letter of Aristeas

Our primary literary source for the story of the Seventy is a Hellenistic work known as the Letter of Aristeas (see Chapter 8). All subsequent traditions regarding the origins of the Septuagint would appear to have stemmed from it (Wasserstein and Wasserstein 2006: 18). Preserved by Christian scribes, the Letter is extant in twenty-three Greek manuscripts (Wright 2015: 30). While its historical veracity has been challenged, it remains an invaluable document.

In a purported letter from Aristeas, a courtier of Ptolemy Philadelphus, to his ‘brother’ Philocrates, we learn that the royal librarian wished to secure a translation of the Jewish Law for the library in Alexandria (Let. Aris. §§9–11). An embassy was dispatched by Philadelphus to the high priest in Jerusalem requesting six learned elders from each of the twelve tribes (§39). The seventy-two elders duly arrived in Alexandria with a copy of the Law (§173). After attending banquets over seven successive days, they retired to the island of Pharos, where they completed the translation in seventy-two days (§307). It was received enthusiastically by the local Jewish community, and a curse was pronounced upon anyone who dared alter it by addition, transposition, or omission (§310–11).

The authenticity of the Letter was generally taken for granted until the early modern period. The earliest registration of doubt was made in passing by the Spanish humanist Juan Louis Vives (1492–1540) in his commentary on Augustine’s City of God (Wasserstein and Wasserstein 2006: 241). Azariah dei Rossi (1513–78), a learned Jewish physician, devoted an entire monograph to the Letter, drawing upon both Jewish and Christian sources, and thus preparing the way for a critical examination of the text (Wasserstein and Wasserstein 2006: 247). It was, however, the publication of Humphrey Hody’s devastating Contra historiam Aristeae de LXX interpretibus dissertatio in 1684 that sealed its fate. Since Hody it has been agreed by most scholars that the document is a fiction, its author in all likelihood a Greek-speaking Jew living in Egypt sometime during the late second century bce. There are indications that he was elaborating upon a local tradition, though whether it counted seventy or seventy-two translators is unknown. The Jewish historian Josephus, writing in the late first century ce, uses both interchangeably (Ant. 12.57). The figure seventy, which prevailed, signals a typological relationship between the birth of the Septuagint and the giving of the Law at Sinai—a ‘constitutional fiction’, as Jellicoe (1968: 57) happily put it, which invests the translation with both sacred and secular authority (Honigman 2007: 142) ...

Lagarde and Kahle

If contemporary specialists do not, as a rule, subscribe to the Aristeas legend, they too assign value to ancient texts, and must likewise trade in theories of origin. For much of the twentieth century two such theories polarized Septuagint scholarship. Both were stimulated by problems in textual criticism, and both addressed the question of which text the Letter of Aristeas was attempting to legitimate (Fernández Marcos 2000: 53).

Each conceptualizes the Septuagint very differently. The Urtext theory, associated with German polymath Paul de Lagarde (1863), posits a single autograph for the Greek Pentateuch, produced sometime in the third century bce, a lineal descendent of which was defended by Aristeas. Although the text of the autograph has been transmitted through continuous copying, every existing manuscript contains material conflated from later editions (Wevers 1985: 21).

The assumption of an Urtext remained uncontested until 1915 when it was challenged in an article by Paul Kahle (Jellicoe 1968: 61). According to Kahle, the received text of the Greek Pentateuch represents a late first-century bce Jewish recension derived from earlier partial translations (or Targums) that had been intended to accompany the public reading of the Hebrew Law. It was the legitimacy of this later standardized edition that was at issue for the author of Aristeas. In support of what became known as the Greek Targum theory, Kahle (1959: 209–64) drew attention to a wide range of textual witnesses, which, on his analysis, did not fit Lagarde’s binary model of an original text subject to interference from later recensions.

Persuasive though it was, Kahle’s argument was ultimately rejected by most specialists. Quite simply, the hypothesis of common descent from a single archetype accounts more economically for the evidence. The current consensus holds that the autograph of each book of the Greek Pentateuch was a unique once-off undertaking by translators who did not substantially revise their work (Tov 2010: 16). Moreover analysis of translation technique points to marked divergences in method between each of the five books, a finding which militates against Kahle’s contention that the text underwent standardization (Wevers 1985: 20). Yet if Lagardian assumptions have prevailed, they have also been significantly qualified. Critical editors have been forced to reckon with a textual situation far more complex than that envisioned by Lagarde (Wevers 1988: 27).

Recent study of the Greek scriptural fragments found at Qumran, particularly 4Q119, containing Leviticus ch. 26 (dating to the late second or first century bce), has reopened the question of textual fluidity (Faulkenberry Miller 2007: 27). To what extent the Urtext theory will be further qualified remains to be seen.

Sitz im Leben

One merit of Kahle’s account is that it explains how the Greek Pentateuch was initially used, namely, as a Targum on its Hebrew source. It thus locates the social impetus for the translation in a liturgical context. A variation on this Sitz im Leben was proposed by Thackeray (1921: 12), who characterized the Greek text as ‘a people’s book designed undoubtedly for synagogue use’. Despite its intuitive appeal, the liturgical theory of Septuagint origins is vulnerable to the charge of anachronism. As Bickerman (1959: 7) pointed out, a regular cycle of Pentateuchal readings is not attested until the second century ce. Nevertheless there are hints of a trend towards liturgical standardization in the writings of the Jewish philosopher Philo (c.25 bce–40 ce) which might be enlisted in support of Thackeray. Cohen (2007: 68) has found that of the twelve distinct quotations of the Latter Prophets made by Philo, at least nine are represented in the string of Haftarot (passages from the prophets that accompany the Torah reading) currently appointed for the liturgical cycle from the 17th of Tammuz to the Day of Atonement. While the statistical significance of this correlation is difficult to assess, it remains highly suggestive ...

The discussion of Sitz im Leben was advanced by Brock (1972) in a seminal paper which suggested that the Greek Pentateuch had played a role similar to that of the Iliad in Hellenistic education. In Brock’s memorable phrase, it was ‘the Homer of the Jews’. This idea has enjoyed renewed attention since the publication of A New English Translation of the Septuagint (Pietersma and Wright 2007). According to the so-called ‘interlinear model’ adopted by the editors, the Greek Pentateuch was originally conceptualized along the lines of the vernacular translations of Homer used as an aid to reading (Pietersma 2002: 346–50). A signal advantage of this proposal is that it speaks to the linguistic evidence. Although the language of the Greek Pentateuch is firmly rooted in the contemporary Koine, it is characterized by varying degrees of formal equivalence (isomorphism) with its Hebrew source.

On the hypothesis that the translation was in some sense a school text, this phenomenon is attributable to its ancillary role with respect to the Hebrew. Whether the verbal makeup of the Greek text is such as to war rant the inference of interlinearity has, however, been vigorously debated. It is of course not difficult to cite instances where isomorphism has trumped idiomatic expression. Yet the detailed findings reported by Lee (2018) must also be taken into account, as they underscore the relative acceptability of much Pentateuchal Greek.

In the absence of secure knowledge regarding contemporary Jewish social practices in Alexandria—whether liturgical, judicial, or educational—all theories of Septuagint origins must remain somewhat speculative ...

---

It is generally assumed that by the first century ce a corpus of Jewish Scriptures in Greek was taking shape in Alexandria parallel to the formation of the Hebrew Bible in Palestine. Some scholars thus refer to the Alexandrian Bible, though such a conception of the matter perhaps owes more to the Aristean legend that to sober historiography (Tov 2010: 4)

Hebrew–Greek Translation in Hellenistic Judaism

Subsequent to the Pentateuch, there was a flurry of Hebrew–Greek translation activity, including not only the books of the later rabbinic canon, but numerous others. For most of these the extant manuscript evidence points to a single Hellenistic archetype. There are, however, apparent exceptions. Judges, Esther, Tobit, and Daniel are each extant in two textual forms, and the textual history underlying them remains somewhat uncertain. While Dines (2004: 59) is inclined to favour the idea of textual multiplicity, and sees the so-called double texts as a challenge to the Lagardian model, Fernández Marcos (2000: 103) concludes that we are dealing with distinct recensions of the same underlying translation (see Chapters 13, 20, 22, 26). In some instances the earliest recoverable form of the Greek text points to a Vorlage at considerable remove from MT. Greek Jeremiah, for instance, is shorter than its Hebrew counterpart and differently ordered; it was likely translated from a Hebrew text that preceded the expanded version transmitted by the Masoretes (Fernández Marcos 2000: 81; and Chapter 18).

What is perhaps most striking about the translations of the Hellenistic period is the heterogeneity of method and style to which they attest, suggestive, as it is, of a diversity of origins and purposes. Thus while the Greek Psalter adheres closely to the linguistic form of its Hebrew source (see Chapter 23), the language of Greek Proverbs exhibits a remarkable sensitivity to stylistic and rhetorical effects (Aitken and Cuppi 2015: 347; see Chapter 24). Some of the translations differ significantly from their Hebrew counterparts with respect to content. Greek Esther, for instance, is more than twice as long as the extant Hebrew version, probably due to novelistic expansion (Hengel 2004: 87; see Chapter 22). The Old Greek of Job, on the other hand, would appear to have abbreviated its source by as much as one sixth (at least in its original form); verses are passed over sometimes two or three at a time (Cox 2015: 386; see Chapter 25). The Hellenistic translations also vary widely in the amount of interpretation or actualization evident to modern readers ...