GREEK VERSIONS OF OT
I. The Septuagint (LXX
). 1. The Septuagint, or Version of the Seventy, has special characteristics which differentiate it strongly from all other versions of the Scriptures. Not only are its relations to the original Hebrew of the OT more difficult and obscure than those of any other version to its original, but, as the Greek OT of the Christian community from its earliest days, it has a special historical importance which no other version can claim, and only the Vulgate can approach. Its history, moreover, is very obscure, and its criticism bristles with difficulties, for the removal of which much work is still needed. The present article can aim only at stating the principal questions which arise in relation to it, and the provisional conclusions at which the leading students of the subject have arrived.
2. There is no doubt that the LXX
originated in Alexandria, in the time of the Macedonian dynasty in Egypt. Greeks had been sporadically present in Egypt even before the conquest of the country by Alexander, and under the Ptolemys they increased and multiplied greatly. Hundreds of documents discovered in Egypt within the last few years testify to the presence of Greeks and the wide-spread knowledge of the Greek language from the days of Ptolemy Soter onwards. Among them, especially in Alexandria, were many Jews, to whom Greek became the language of daily life, while the knowledge of Aramaic, and still more of literary Hebrew, decayed among them. It was among such surroundings that the LXX
came into existence. The principal authority on the subject of its origin is the Letter of Aristeas (edited by H. St. J. Thackeray in Swete’s Introduction to the OT in Greek
, and by P. Wendland in the Teubner series
). This document, which purports to be written by a Greek official of high rank in the court of Ptolemy ii. (Philadelphus, b.c. 285 247), describes how the king, at the suggestion of his librarian, Demetrius of Phalerum, resolved to obtain a Greek translation of the laws of the Jews for the library of Alexandria; how, at the instigation of Aristeas, he released the Jewish captives in his kingdom, to the number of some 100,000, paying the (absurdly small) sum of 20 drachmas apiece for them to their masters; how he then sent presents to Eleazar, the high priest at Jerusalem, and begged him to send six elders out of each tribe to translate the Law; how the 72 elders were sent, and magnificently entertained by Ptolemy, and were then set down to their work in the island of Pharos; and how in 72 days they completed the task assigned to them. The story is repeated by Josephus ( Ant. XII. ii.) from Aristeas in a condensed form. In later times it received various accretions, increasing the miraculous character of the work; but these additions have no authority.
3. That the Letter of Aristeas is substantially right in assigning the original translation of the Law to the time of one of the early Ptolemys there is no reason to doubt; but the story has the air of having been considerably written up, and it is impossible to say precisely where history stops and fiction begins. Demetrius of Phalerum was librarian to Ptolemy i., but was in disgrace under his successor, and died about 283; hence he can hardly have been the prime mover in the affair. But if not, the writer of the Letter cannot have been the person of rank in Ptolemy’s court that he represents himself to be, and the credit of the document is severely shaken. It cannot be depended on for accuracy in details, and it is necessary to turn to the internal evidence for further information. It will be observed that Aristeas speaks only of ‘the Law,’ i.e. the Pentateuch; and there is no reason to doubt that this was the first part of the OT to be translated, and that the other books followed at different times and from the hands of different translators. A lower limit for the completion of the work, or of the main part of it, is given in the prologue to Sirach (written probably in b.c. 132), where the writer speaks of ‘the law itself and the prophets and the rest of the books’ ( sc . the Hagiographa) as having been already translated. It may therefore be taken as fairly certain that the LXX
as a whole was produced between b.c. 285 and 150.
4. Its character cannot be described in a word. It is written in Greek, which in vocabulary and accidence is substantially that koinç dialektos , or Hellenistic Greek, which was in common use throughout the empire of Alexander, and of which our knowledge, in its non-literary form, has been greatly extended by the recent discoveries of Greek papyri in Egypt. In its syntax, however, it is strongly tinged with Hebraisms, which give it a distinct character of its own. The general tendency of the LXX
translators was to be very literal, and they have repeatedly followed Hebrew usage (notably in the use of pronouns, prepositions, and participial constructions) to an extent which runs entirely counter to the genius of the Greek language. [For examples, and for the grammar of the LXX
generally, see the Introduction to Selections from the Septuagint , by F. C. Conybeare and St. George Stock (1905).] The quality of the translation differs in different books. It is at its best in the Pentateuch, which was probably both the first and the most deliberately prepared portion of the translation. It is at its worst in the Prophets, which presented the greatest difficulties in the way of interpretation. Neither the Greek nor the Hebrew scholarship of the translators was of a high order, and they not infrequently wrote down words which convey no rational meaning whatever. Something has been done of late to distinguish the work of different translators.
iv. 245, 398, 578, viii. 262, the results of which are here summarized.] It has been shown that Jer. is probably the work of two translators, who respectively translated chs. 1 28 and 29 51 (in the Greek order of the chapters), the latter, who was an inferior scholar, being responsible also for Baruch. Ezek. likewise shows traces of two translators, one taking chs. 1 27 and 40 48, the other 28 39. The Minor Prophets form a single group, which has considerable affinities with the first translators of both Jer. and Ezekiel. Isaiah stands markedly apart from all these, exhibiting a more classical style, but less fidelity to the Hebrew. 1Kings (= 1Samam.) similarly stands apart from 2 4 Kings, the latter having features in common with Judges.
5. Some other features of the LXX
must be mentioned which show that each book, or group of books, requires separate study. In Judges the two principal MSS (Codd. A and B, see below, § 10) differ so extensively as to show that they represent different recensions. In some books (notably the latter chapters of Exodus 3:1-22
K 4 11, Proverbs 24:1-34
; Proverbs 25:1-28
; Proverbs 26:1-28
; Proverbs 27:1-27
; Proverbs 28:1-28
; Proverbs 29:1-27
, Jeremiah 25:1-38
; Jeremiah 26:1-24
; Jeremiah 27:1-22
; Jeremiah 28:1-17
; Jeremiah 29:1-32
; Jeremiah 30:1-24
; Jeremiah 31:1-40
; Jeremiah 32:1-44
; Jeremiah 33:1-26
; Jeremiah 34:1-22
; Jeremiah 35:1-19
; Jeremiah 36:1-32
; Jeremiah 37:1-21
; Jeremiah 38:1-28
; Jeremiah 39:1-18
; Jeremiah 40:1-16
; Jeremiah 41:1-18
; Jeremiah 42:1-22
; Jeremiah 43:1-13
; Jeremiah 44:1-30
; Jeremiah 45:1-5
; Jeremiah 46:1-28
; Jeremiah 47:1-7
; Jeremiah 48:1-47
; Jeremiah 49:1-39
; Jeremiah 50:1-46
; Jeremiah 51:1-64
) the order of the LXX
differs completely from that of the Hebrew, testifying to an arrangement of the text quite different from that of the Massoretes. Elsewhere the differences are not in arrangement but in contents. This is especially the case in the latter chapters of Jos.
, 1Kings (= 1Samam.) 17 18, where the LXX
omits (or the Heb. adds) several verses; 3 K 8 and 12, where the LXX
incorporates material from some fresh source; Psa 151:1-7
, which is added in the LXX
; Job, the original LXX
text of which was much shorter than that of the Massoretic Hebrew; Esther, where the Greek has large additions, which now appear separately in our Apocrypha, but which are an integral part of the LXX
; Jer., where small omissions and additions are frequent; and Daniel, where the LXX
includes the episodes of Susanna, Bel and the Dragon, and the Song of the Three Children, which have now been relegated (in obedience to Jerome’s example) to the Apocrypha.
6. The mention of the Apocrypha suggests the largest and most striking difference between the LXX
and the Hebrew OT, namely, in the books included in their respective canons; for the Apocrypha, as it stands to-day in our Bibles, consists (with the exception of 2 Esdras and the Prayer of Manasseh) of books which form an integral part of the LXX
canon, but were excluded from the Hebrew canon when that was finally determined about the end of the 1st century
. Nor did these books stand apart from the others in the LXX
as a separate group. The historical books (1 Esdras, Tob., Judith, and sometimes Mac.) have their place with Chron., Ezr., Neh.; the poetical books (Wisd., Sir.) stand beside Prov., Eccles., and Cant.; and Baruch is attached to Jeremiah. The whole arrangement of the OT books differs, indeed, from the stereotyped order of the Massoretic Hebrew. The latter has its three fixed divisions (i) the Law, i.e. the Pentateuch; (ii) the Prophets, consisting of the Former Prophets (Jos.
, Judges 1:1-36
; Judges 2:1-23
; Judges 3:1-31
; Judges 4:1-24
Kings) and the Latter Prophets (Isaiah, Jeremiah, Ezekiel, and the Minor Prophets); (iii) the Hagiographa, including Chron., Ps., Job, Prov., Ruth, Cant., Eccles., Lam., Esth., Dan., Ezr., Nehemiah. But the LXX
attaches Ruth to Judges, Chron. and Ezr.-Neh. to Kings, Baruch and Lam. to Jer., and Dan. to the three Greater Prophets. Its principle of arrangement is, in fact, different. In place of divisions which substantially represent three different stages of canonization, it classifies the books in groups according to the character of their subject-matter Law, History, Poetry, and Prophecy. The details of the order of the books differ in different MSS and authoritative lists, but substantially the principle is as here stated; and the divergence has had considerable historical importance. In spite of the dissent of several of the leading Fathers, such as Origen and Athanasius, the LXX
canon was generally accepted by the early Christian Church. Through the medium of the Old Latin Version it passed into the West, and in spite of Jerome’s adoption of the Hebrew canon in his Vulgate, the impugned books made their way back into all Latin Bibles, and have remained there from that day to this.
vii. 343 (1906).] In the Reformed Churches their fate has been different; for the German and English translators followed Jerome in adopting the Hebrew canon, and relegated the remaining books to the limbo of the Apocrypha. The authority attaching to the LXX
and Massoretic canons respectively is a matter of controversy which cannot be settled offhand; but the fact of their divergence is certain and historically important.
7. If the LXX
had come down to us in the state in which it was at the time when its canon was complete (say in the 1st cent. b.c.), it would still have presented to the critic problems more than enough, by reason of its differences from the Hebrew in contents and arrangement, and the doubt attaching to its fidelity as a translation; but these difficulties are multiplied tenfold by the modifications which it underwent between this time and the date to which our earliest MSS belong (4th cent. a.d.). It has been shown above that the LXX
was the Bible of the Greek-speaking world at the time when Christianity spread over it. It was in that form that the Gentile Christians received the OT; and they were under no temptation to desert it for the Hebrew Bible (which was the property of their enemies, the Jews), even if they had been able to read it. The LXX
consequently became the Bible of the early Christian Church, to which the books of the NT were added in course of time. But the more the Christians were attached to the LXX
, the less willing became the Jews to admit its authority; and from the time of the activity of the Rabbinical school of Jamnia, about the end of the 1st cent., to which period the fixing of the Massoretic canon and text may be assigned with fair certainty, they definitely repudiated it. This repudiation did not, however, do away with the need which non-Palestinian Jews felt for a Greek OT; and the result was the production, in the course of the 2nd cent., of no less than three new translations. These translations, which are known under the names of Aquila, Theodotion, and Symmachus, are described below (§§ 15 18 ); here it is sufficient to say that they were all translated from the Massoretic OT, and represent it with different degrees of fidelity, from the pedantic verbal imitation of Aquila to the literary freedom of Symmachus. By the beginning of the 3rd cent. there were, therefore, four Greek versions of the OT in the field, besides portions of others which will be mentioned below.
8. Such was the state of things when Origen (a.d. 185 253), the greatest scholar produced by the early Church, entered the field of textual criticism. His labours therein had the most far-reaching effect on the fortunes of the LXX
, and are the cause of a large part of our difficulties in respect of its text to-day. Struck by the discrepancies between the LXX
and the Heb., he conceived the idea of a vast work which should set the facts plainly before the student. This was the Hexapla , or sixfold version of the OT, in which six versions were set forth in six parallel columns. The six versions were as follows (1) the Hebrew text; (2) the same transliterated in Greek characters; (3) the version of Aquila, which of all the versions was the nearest to the Hebrew; (4) the version of Symmachus; (5) his own edition of the LXX
; (6) the version of Theodotion. In the case of the Psalms, no less than three additional Greek versions were included, of which very little is known; they are called simply Quinta, Sexta , and Septima . Elsewhere also there is occasional evidence of an additional version having been included; but these are unimportant. A separate copy of the four main Greek versions was also made, and was known as the Tetrapla. The principal extant fragment of a MS of the Hexapla (a 10th cent. palimpsest at Milan, containing about 11 Psalms) omits the Hebrew column, but makes up the total of six by a column containing various isolated readings. The only other fragment is a 7th cent. leaf discovered at Cairo in a genizah (or receptacle for damaged and disused synagogue MSS), and now at Cambridge. It contains Psalms 22:15-18
; Psalms 22:20-28
, and has been edited by Dr. C. Taylor ( Cairo Genizah Palimpsests , 1900).
Origen’s Hebrew text was substantially identical with the Massoretic; and Aq.
, Symm., and Theod., as has been stated above, were translations from it; but the LXX
, in view of its wide and frequent discrepancies, received special treatment. Passages present in the LXX
, but wanting in the Heb., were marked with an obelus ( or ); Passages Wanting In The LXX
, But Present In The Heb., Were Supplied From Aq.
Or Theod., And Marked With An Asterisk (*); The Close Of The Passage To Which The Signs Applied Being Marked By A Metobelus (: Or %. Or ×). In Cases Of Divergences In Arrangement, The Order Of The Heb. Was Followed (Except In Prov.), And The Text Of The LXX
Was Considerably Corrected So As To Bring It Into Better Conformity With The Heb. The Establishment Of Such A Conformity Was In Fact Origen’s Main Object, Though His Conscience As A Scholar And His Reverence For The LXX
Did Not Allow Him Altogether To Cast Out Passages Which Occurred In It, Even Though They Had No Sanction In The Hebrew Text As He Knew It.
9. The great MSS of the Hexapla and Tetrapla were preserved for a long time in the library established by Origen’s disciple, Pamphilus, at Cæsarea, and references are made to them in the scholia and subscriptions of some of the extant MSS of the LXX
(notably א and Q). So long as they were in existence, with their apparatus of critical signs, the work of Origen in confusing the Gr. and Heb. texts of the OT could always be undone, and the original texts of the LXX
substantially restored. But MSS so huge could not easily be copied, and the natural tendency was to excerpt the LXX
column by itself, as representing a Greek text improved by restoration to more authentic form. Such an edition, containing Origen’s fifth column, with its apparatus of critical signs, was produced early in the 4th cent. by Pamphilus, the founder of the library at Cæsarea, and his disciple Eusebius; and almost simultaneously two fresh editions of the LXX
were published in the two principal provinces of Greek Christianity, by Hesychius at Alexandria, and by Lucian at Antioch. It is from these three editions that the majority of the extant MSS of the LXX
have descended; but the intricacies of the descent are indescribably great. In the case of Hexaplaric MSS, the inevitable tendency of scribes was to omit, more or less completely, the critical signs which distinguished the true LXX
text from the passages imported from Aq.
or Theod.; the versions of Aq.
, Theod., and Symm. have disappeared, and exist now only in fragments, so that we cannot distinguish all such interpolations with certainty; Hexaplaric, Hesychian, and Lucianic MSS acted and reacted on one another, so that it is very difficult to identify MSS as containing one or other of these editions; and although some MSS can be assigned to one or other of them with fair confidence, the majority contain mixed and undetermined texts. The task of the textual critic who would get behind all this confusion of versions and recensions is consequently very hard, and the problem has as yet by no means been completely solved.
10. The materials for its solution are, as in the NT, threefold Manuscripts, Versions, Patristic Quotations; and these must be briefly described. The earliest MSS are fragments on papyrus, some of which go back to the 3rd century. About 16 in all are at present known, the most important being (i) Oxyrhynchus Pap. 656 (early 3rd cent.), containing parts of Genesis 14:1-24
; Genesis 15:1-21
; Genesis 16:1-16
; Genesis 17:1-27
; Genesis 18:1-33
; Genesis 19:1-38
; Genesis 20:1-18
; Genesis 21:1-34
; Genesis 22:1-24
; Genesis 23:1-20
; Genesis 24:1-67
; Genesis 25:1-34
; Genesis 26:1-35
; Genesis 27:1-46
, where most of the great vellum MSS are defective; (ii) Brit. Mus. Pap. 37 (7th cent.), sometimes known as U, containing the greater part of Psalms 10:1-18
; Psalms 11:1-7
; Psalms 12:1-8
; Psalms 13:1-6
; Psalms 14:1-7
; Psalms 15:1-5
; Psalms 16:1-11
; Psalms 17:1-15
; Psalms 18:1-50
; Psalms 19:1-14
; Psalms 20:1-9
; Psalms 21:1-13
; Psalms 22:1-31
; Psalms 23:1-6
; Psalms 24:1-10
; Psalms 25:1-22
; Psalms 26:1-12
; Psalms 27:1-14
; Psalms 28:1-9
; Psalms 29:1-11
; Psalms 30:1-12
; Psalms 31:1-24
; Psalms 32:1-11
; Psalms 33:1-22
; Psalms 34:1-22
4; (iii) a Leipzig papyrus (4th cent.), containing Psalms 30:1-12
; Psalms 31:1-24
; Psalms 32:1-11
; Psalms 33:1-22
; Psalms 34:1-22
; Psalms 35:1-28
; Psalms 36:1-12
; Psalms 37:1-40
; Psalms 38:1-22
; Psalms 39:1-13
; Psalms 40:1-17
; Psalms 41:1-13
; Psalms 42:1-11
; Psalms 43:1-5
; Psalms 44:1-26
; Psalms 45:1-17
; Psalms 46:1-11
; Psalms 47:1-9
; Psalms 48:1-14
; Psalms 49:1-20
; Psalms 50:1-23
; Psalms 51:1-19
; Psalms 52:1-9
; Psalms 53:1-6
; Psalms 54:1-7
; Psalms 55:1-23
, the first five being considerably mutilated; (iv) a papyrus at Heidelberg (7th cent.), containing Zechariah 4:6 Malachi 4:5
. A papyrus at Berlin, containing about two-thirds of Gen., and said to be of the 4th or 5th cent., is not yet published.
The principal vellum uncial MSS, which are of course the main foundation of our textual knowledge, are as follows. See also Text of NT.
א or S. Codex Sinaiticus , 4th cent., 43 leaves at Leipzig, 156 (besides the whole NT) at St. Petersburg, containing fragments of Geo. and Num., 1 Chronicles 9:27
to 1 Chronicles 19:17
, 2E Esther 9:9
to end, Esth., Tob., Jdt 1:1-16
and 4 Mac., Is., Jer., Lamentations 1:1
to Lamentations 2:20
, Joel, Obad., Jon., Nah. Mal., and the poetical books. Its text is of a very mixed character. It has a strong element in common with B, and yet is often independent of it. In Tob. it has a quite different text from that of A and B, and is perhaps nearer to the original Heb. Its origin is probably composite, so that it is not possible to assign it to any one school. Its most important correctors are Can and C b , both of the 7th cent., the former of whom states, in a note appended to Esth., that he collated the MS with a very early copy, which itself had been corrected by the hand of Pamphilus.
A. Codex Alexandrinus , 5th cent., in the British Museum; complete except in Psalms 49:19
to Psalms 79:10
and smaller lacunæ, chiefly in Genesis 3:1-24
and 4 Mac. are included. The Psalter is liturgical, and is preceded by the Epistle of Athanasius on the Psalter, and the Hypotheseis of Eusebius; the Canticles are appended to it. The text is written by at least two scribes; the principal corrections are by the original scribes and a reviser of not much later date. It is almost certainly of Egyptian origin, and has sometimes been supposed to represent the edition of Hesychius, but this is by no means certain yet. In Judges it has a text wholly different from that of B, and in general the two MSS represent different types of text; the quotations from the LXX
in the NT tend to support A rather than B.
B. Codex Vaticanus , 4th cent., in the Vatican; complete, except for the loss of Genesis 1:1
to Genesis 46:28
, 2Ki 2:5-7
; 2 Kings 2:10-13
, Psalms 105:27
to Psalms 137:6
, and the omission of 1 4 Maccabees. Its character appears to differ in different books, but in general Hort’s description seems sound, that it is closely akin to the text which Origen had before him when he set about his Hexapla. It is thus of Egyptian origin, and is very frequently in accord with the Bohairic version. Recently Rahlfs has argued that in Ps. it represents the edition of Hesychius, but his proof is very incomplete; for since he admits that Hesychius must have made but few alterations in the pre-Origenian Psalter, and that the text of B is not quite identical with that which he takes as the standard of Hesychius (namely, the quotations in Cyril of Alexandria), his hypothesis does not seem to cover the phenomena so well as Hort’s. The true character of B, however, still requires investigation, and each of the principal groups of books must be examined separately.
C. Codex Ephræmi rescriptus , 5th cent., at Paris; 64 leaves palimpsest, containing parts of the poetical books.
D. The Cotton