The Septuagint (Gr. oa 0', Lat. LXX.), or the "Alexandrian version of the Old Testament," so named from the legend of its composition by seventy (Lat. septuaginta), or more exactly seventy-two, translators. In the Letter of Aristeas to Philocrates 1 this legend is recounted as follows: Demetrius of Phalerum, keeper of the Alexandrian library, proposed to King Ptolemy II. Philadelphus (285-247 B.C.) to have a Greek translation of the Jewish law made for the library. The king consented and, after releasing 10o,000 Jewish captives in his kingdom, sent an embassy with rich presents to the high priest Eleazar at Jerusalem asking him to send six ancient, worthy and learned men from each of the twelve tribes to translate the law for him at Alexandria. Eleazar readily sent the seventy-two men with a precious 1 Edited by H. St J. Thackeray in H. B. Swete's Introd. to the Old Testament in Greek (1900), and by P. Wendland in the Teubner series (1900).
roll of the law. They were honourably received at the court of Alexandria and conducted to the island (Pharos), that they might work undisturbed and isolated. When they had come to an agreement upon a section Demetrius wrote down their version; the whole translation was finished in seventy-two days. The Jewish community of Alexandria was allowed to have a copy, and accepted the version officially; indeed a curse was laid upon the introduction of any changes in it.
There is no question that this Letter (which is condensed in Josephus, Ant. xii. 2) is spurious.' Aristeas, an official at Ptolemy's court, is represented as a heathen, but the real writer must have been a Jew and no heathen. Aristeas is represented as himself a member of the embassy to Eleazar; but the author of the Letter cannot have been a contemporary of the events he records, else he would have known that Demetrius fell out of favour at the very beginning of the reign of Philadelphus, on a charge of intriguing against his succession to the throne. 2 Nor could a genuine honest witness have fallen into the absurd mistake of making delegates from Jerusalem the authors of the Alexandrian version. There are also one or two passages (§§ 28, 182) where the author seems to forget that he is playing the role of Aristeas. The forgery, however, seems to be an early one. 3 "There is not a court-title, an institution, a law, a magistracy, an office, a technical term, a formula, a peculiar phrase in this letter which is not found on papyri or inscriptions and confirmed by them." 4 That in itself would not necessarily imply a very early date for the piece; but what is decisive is that the author limits canonicity to the law and knows of no other holy book already translated into Greek. Nor does he claim any inspiration for the translators. Further, what he tells about Judaea and Jerusalem is throughout applicable to the period when the Ptolemies bore sway there and gives not the slightest suggestion of the immense changes that followed the conquest of Palestine by the Seleucids. It is probable that the Jewish philosopher Aristobulus, who lived under Ptolemy VI. Philometor (180-145 B.C.), derived his account of the origin of the LXX. from this Letter, with which it corresponds.' There seems good ground for believing that the letter contains some elements derived from actual tradition as to the origin of the LXX. Ptolemy Philadelphus was a king of eclectic literary tastes, and the welcome he gave to a Buddhist mission from India might well have been extended to Jews from Palestine. The letter lays great stress on the point that the LXX. is the official and authoritative Bible of the Hellenistic Jews, having not only been formally accepted by the synagogue at Alexandria, but authorized by the authorities at Jerusalem. This, and the fact that the style of the version is not that of a book intended for literary use, points to the conclusion that the translation was made to satisfy the religious needs of the Jews in Alexandria, and possibly also in the hope of gaining proselytes. In view of the Jewish prejudice against writing Scripture in any but the old holy form (the Targum, for instance, was for centuries handed down orally), it is quite possible that some impulse to the Alexandrian version came from without. Philadelphus may have encouraged it both to satisfy his own curiosity and to promote the use of G leek among the large Jewish population of the city. That the work is purely Jewish in character is 1 Its claims were demolished by Humphry Hody, Regius Professor of Greek at Oxford, in 1684.
2 Hermippus Callimachius, ap. Diog. Laert. v. 78. Irenaeus indeed, evidently following some other account, fixes the translation in the time of Ptolemy I.
P. Wendland, however, puts it after the Maccabean age (say 96 B.C.) and before the Roman invasion of Palestine (63,B.C.). ' G. Lumbroso, Recherches sur l'econ. pol. de l'Egypte sous les Lagides (Turin, 1870), p. xiii.
5 Clem. Alex. Strom. i. p. 342, ed. Sylb.; Eusebius, Praep. Ev. ix. 6, p. 420 seq.; cf. Valckenaer, Diatribe de Aristobulo (Leiden, 1806), reprinted in Gaisford's edition of the Praep. Ev. One must not overlook the possibility that Aristobulus's Interpretation of the Holy Laws may itself be the pseudonymous work of some otherwise unknown Jewish author. It and the Letter of Aristeas seem to be of the same date, if not even by the same hand. And Philo (Vita Mosis, ii. § 7, ii. 141) describes an annual festival held at Pharos in honour of the origin of the Greek Bible.
only what was inevitable in any case. The translators were necessarily Jews, though Egyptian and not Palestinian Jews, and were necessarily and entirely guided by the living tradition which had its focus in the synagogal lessons.' And hence it is easily understood that the version was ignored by the Greeks, who must have found it barbarous and largely unintelligible, but obtained speedy acceptance with the Jews, first in private use and at length also in the synagogue service.
The next direct evidence which we have as to the origin of the LXX. is the prologue to Ecclesiasticus, from which it appears that about 130 B.C. not only the law but "the prophets and the other books" were extant in Greek.' With this it agrees that the text of Ecclesiasticus and the other ancient relics of JewishGreek literature, preserved in the extracts made by Alexander Polyhistor (Eusebius, Praep. Ev. ix.), all show acquaintance with the LXX. 8 The experiment on the Pentateuch (of which alone Aristeas speaks) had evidently been extended to other rolls as they arrived from Jerusalem. These later translations were not made simply to meet the needs of the synagogue, but express a literary movement among the Hellenistic Jews, stimulated by the favourable reception given to the Greek Pentateuch, which enabled the translators to count on finding an interested public. If a translation was well received by reading circles among the Jews, it gradually acquired public acknowledgment and was finally used also in the synagogue, so far as lessons from other books than the Pentateuch were used at all. But originally the translations were mere private enterprises, as appears from the prologue to Ecclesiasticus and the colophon to Esther. It appears also that it was long before the whole Septuagint was finished and treated as a complete work. We may grant that the Pentateuch (and perhaps part of Joshua) was translated in the 3rd century B.C. The other books followed, generally speaking, in the order in which they occur in the Hebrew Canon. Isaiah perhaps dates from c. 80, Jeremiah, Ezekiel and the Twelve Prophets, as also r Kings (=1 Samuel), c. r 50. Most of the "Writings," together with Judges and 2-4 Kings, were probably translated in the 1st century B.C., while Ecclesiastes and Daniel (the latter incorporated from Theodotion) date only from the 2nd century of the Christian era. As the work of translation went on so gradually, and new books were always added to the collection, the compass of the Greek Bible came to be somewhat indefinite. The law always maintained its pre-eminence as the basis of the canon; but the prophetic collection changed its aspect by having various Hagiographa incorporated with it according to an arbitrary arrangement by subjects. The distinction made in Palestine between Hagiographa and Apocrypha was never properly established among the Hellenists. In some books the translators took the liberty of making considerable additions to the original, e.g. those to Daniel, and these additions became a part of the Septuagint. Nevertheless, learned Hellenists were quite well aware of the limits of the canon and respected them. Philo can be shown to have known the Apocrypha, but he never cites them, much less allegorizes them or uses them in proof of his tenets. And in some measure the widening of the Old Testament canon in the Septuagint must be laid to the account of Christians.
The vocabulary and accidence of the Greek of the Septuagint are substantially those of the ' cow?) SLhXEKros or Hellenistic Greek spoken throughout the empire of Alexander. The language of the Pentateuch attains the higher level shown by the papyri of the early Ptolemaic age, that of the prophets reflects the less literary style of the papyri of c. 130-100 B.C. In the latest parts of the translation Mr St John Thackeray notes two opposing influences, (a) the growing reverence for the letter of Scripture, tending to a pedantic literalism, (b) the influence of the Atticistic school, strongest in free writings like 4 Maccabees but leaving its mark also on 4 Kings. But if in some respects the Septuagint is the great monument of the Koo* in e It is quite likely that they worked on rolls newly brought from Jerusalem. There was no desire to found an Alexandrian canon or type of text.
This does not necessarily mean that the whole of the section of the Hebrew Old Testament known as "The Writings" was translated by that date.
8 Philo seems to have known the Greek version of most of the Old Testament except Esther, Ecclesiastes, Canticles and Daniel.
others, especially in syntax, it is strongly tinged with Hebraisms, and there are many passages where it is difficult, if not impossible, to extract any rational meaning. In some cases a book bears the marks of two hands: thus Jeremiah i.-xxviii. was not translated by the worker that undertook ch. xxix.-li. (the former is indifferent, the latter unintelligible Greek), and in Ezekiel one hand is responsible for ch. i.-xxvii., xl.-xlviii., and another for ch. xxviii.-xxxix. (except xxxvi. 24-38). So I Kings stands apart from 2-4 Kings. Isaiah is more akin to classical Greek; like the Pentateuch and I Maccabees it is good icoivi 7. The two chief MSS. of Judges vary so much as to point to different recensions. In some books, especially Jeremiah xxv.-li., the order of the Septuagint is totally different from that of the Massoretic Hebrew text (cf. also Proverbs xxiv.-xxix.). In other cases, notably in Job, the original LXX. text was much shorter than that of the Massoretes; in Esther and Daniel there are numerous additions. The Septuagint does not keep the triple Hebrew division of Law, Prophets and Hagiographa or Writings, but instead of this order of canonization principle it groups its books according to subject matter, Law, History, Poetry, Prophecy, a divergence which had much importance for the history of the Old Testament canon in the Christian church. The early Christians generally accepted the LXX. canon, which through the old Latin, despite Jerome's Vulgate adoption of the Hebrew canon, passed into the West, and into the Latin Bibles, where the Apocrypha (except I Esdras) are still included. The German and English churches followed Jerome in giving a less honoured place to the impugned books.
The Septuagint came into general use with the Grecian Jews even in the synagogue. Philo and Josephus use it, and so do the New Testament writers. But at an early date small corrections seem to have been introduced, especially by such Palestinians as had occasion to use the LXX., in consequence partly of divergent interpretation, partly of differences of text or of pronunciation (particularly of proper names). The Old Testament passages cited by authors of the first century of the Christian era, especially those in the Apocalypse, show many such variations from the Septuagint, and, curiously enough, these often correspond with the later versions (particularly with Theodotion), so that the latter seem to rest on a fixed tradition. Corrections in the pronunciation of proper names so as to come closer to the Massoretic pronunciation are especially frequent in Josephus. Finally a reaction against the use of the Septuagint set in among the Jews after the destruction of the temple - a mpvement which was connected with the strict definition of the canon and the fixing of an authoritative text by the rabbins of Palestine. But long usage had made it impossible for the Jews to do without a Greek Bible, and to meet this want a new version was prepared corresponding accurately with the canon and text of the Pharisees. This was the version of Aquila, which took the place of the Septuagint in the synagogues, and long continued in use there. On this, together with the versions of Theodotion and Symmachus, Origen's Hexapla, and the recensions of Hesychius and Lucian, see Bible (Old Testament, " Texts and Versions").
The LXX. is of great importance in more than one respect. "It was the first step towards that fusion of the Hebraic with the Hellenic strain, which has issued in the mind and heart of modern Christendom. Like the opening of tale Suez Canal it let the waters of the East mingle with those of the West, bearing with them many a freight of precious merchandise." Again, it is probably the oldest translation of considerable extent that ever was written, and at any rate it is the starting-point for the history of Jewish interpretation and the Jewish view of Scripture. And from this its importance as a document of exegetical tradition, especially in lexical matters, may be easily understood. It was in great part com p osed before the close of the canon - nay, before some of the Hagiographa were written - and in it alone are preserved a number of important ancient Jewish books that were not admitted into the canon. As the book which created or at least codified the dialect of Biblical Greek, it is the key to the New Testament and all the literature connected with it. To many its chief value lies in the fact that it is the only independent witness for the text of the Old Testament which we have to compare with the Massoretic text. It may seem that the critical value of the LXX. is greatly impaired, if not entirely cancelled, by the corrupt state of the text. If we have not the version itself in authentic form we cannot reconstruct with certainty the Hebrew text from which it was made, and so cannot get at various readings which can be confidently confronted with the Massoretic text; and it may be a long time before we possess a satisfactory edition of the genuine Septuagint. The difficulties in getting behind the confusion of versions and recensions to produce such a result are indeed formidable. The materials at our disposal are of the usual threefold kind, Manuscripts, Versions and Patristic Quotations. The earliest MSS. are about a score of fragments on papyrus, a few of which go back to the 3rd century A.D. The chief uncial MSS. are, as for the New Testament rc, A, B, C and others. Of these A and B are largely complete, but though both of Egyptian origin vary considerably. A (with which the quotations in the New Testament generally agree) may represent the edition of Hesychius; B, which is often, especially in the Psalms, in accord with the Bohairic version, resembles the text used by Origen in the Hexapla. Of versions the Bohairic (Lower Egypt), the Sahidic (Upper Egypt), the various Syriac translations (unfortunately we have no Old Syriac for the Old Testament), and the Latin (Old Latin and Vulgate, especially the former) are the most important. The evidence of the Fathers is valuable as helping to distinguish local types of text. The testimony of the earliest patristic quotations seems to be in favour of A rather than B. The immediate aim of textual criticism is a recovery of the three main editions, those of Origen, Lucian and Hesychius, and then of the pre-Origenian LXX. text, which lies behind them all. When this has been accomplished there still remains the problem of the relation of the LXX. to the Hebrew. There is no doubt that the Hebrew text from which the LXX. translators worked was often divergent from that represented by the Massoretic. For the Pentateuch we have additional material in the Samaritan version, but here the variants are least. In view of the palpable mistakes made by the Septuagint translators and their often inadequate knowledge of Hebrew, we must not hastily assume that in cases of difference the Greek is to be preferred. The book of Ecclesiasticus (the Hebrew of which has recently been discovered) furnishes a useful lesson here. Yet there is no doubt that much (e.g. in I Samuel) may be learned from the Septuagint; all one can say is that each case must be treated on its own merits.
Editions
. - The Septuagint was first printed in the Complutensian Polyglot (1514-1517), but before it was published in 1521 Aldus published another edition in 1519. The Textus Receptus issued by Pope Sixtus V. (Rome, 1587) was based mainly on Cod. Vaticanus (B) with some collection of the Venice MS. (V). This edition was the basis of the great work of R. Holmes and J. Parsons (Oxford, 1798-1827), who furnished the Sixtine text with an apparatus (not always accurate) drawn from 20 uncials and nearly 280 minuscule MSS., in addition to versions. In1707-1720Grabe had published an edition based on Cod. Alexandrinus (A). C. Tischendorf's text (1850; 7th ed., 1887) was a revision of that of Holmes and Parsons with aft apparatus drawn from the chief uncials. H. B. Swete's edition in 3 vols. (1887-1894; revised 1895-1899) gives the text of B, and, where this fails, that of A or with variant readings from the chief uncials. The larger Cambridge edition, begun in 1906 by A. E. Brooke and N. McLean, follows the same plan with the text, but its apparatus includes all the uncials, the best and most representative minuscules, and the chief versions and patristic quotations.
LITERATURE. - H. B. Swete, Introduction to the Old Testament in Greek (1900); E. Nestle, Septuagintastudien (1886-190); F. G. Kenyon, Our Bible and the Ancient MSS., pp. 48-92 1898); A. Rahlfs, Septuaginta-Studien (1904, Kings; 1907, Psalms); E. Hatch and H. A. Redpath, A Concordance to the Septuagint (Oxford, 1897-1906); H. St J. Thackeray, A Grammar of the Old Testament in Greek, vol. i. (Cambridge, 1909), containing a useful Septuagint bibliography; F. C. Conybeare and St G. Stock, Selections from the Septuagint (Boston and London, 1905); the articles in the various Bible-dictionaries, and other works mentioned in the course of this article. (A. J. G.)
