\.'' YALE DIVINITY SCHOOL LIBRARY Gift of Professor George Dahl THE PEINCIPAL YEESIONS OE BAEUCH BY R E. HAEWELL A DiSSEETATION PeESENTED TO THE FaCULTT OF THE Graduate School op Yale Univeesity IN Candidacy foe the Degeee of Doctoe of Philosophy PEEFACE. The purpose of this investigation of the principal versions of Baruch is to bring forward some new results that have been reached : It is conclusively shown that the whole book was written origi nally in Hebrew. 3 : 9-4 : 4 was a poem in 3/3 meter; 4 : 5-5 : 9, its companion poem, was in 3/2 meter. The Greek version, which was made by a single hand at an early date, has suffered much variation and corruption, especially in the prose part. The present text is late and has probably been influenced much by the recension (not a new translation) bf Theodotion. The Old Syriac version was originally made from the Greek. The only text which is known to us is a later conflate recension; this I have called the Ordinary Syriac version. There is no evidence that it was ever directly influenced by any Hebrew text. Two Latin translations of the book were made. The first of these (Kneucker's Vet. Lat. h) was made from the pre-Hexaplaric Greek. The. second (Kneucker's Vet. Lat. a, surviving chiefly in the Vulgate) is a later close rendering of our standard Greek text. A third text, Hoberg's Die dlteste Lateinische Uehersetzung des Buches Baruch, is the result of conforming Vet. Lat. h to the standard Greek text. It has passed since 1902 as the "oldest Latin version of Baruch", but is really late and unimportant. A pre-Christian date for the Greek translation of Baruch is proved by the identity of its author with that of the Greek of Jeremiah of which it was a part. I wish to express my cordial thanks to Prof. Charles C. Torrey, under whose direction this investigation has been made, for valuable suggestions and assistance. R. R. Harwell. Yale University, April, 1915. TABLE OF CONTENTS PAGE Preface 3 Introduction 4 I. The Syeo-Hexaplae Veesion 5 II. The Oedinaey Syriac Version 10 1. The Syriac Version Translated from the Greek 11 2. Additions to the Syriac Version Based on Greek Authority ] 7 3. Conflations in the Syriac Version Probably Due to a Combination of Variant Greek Readings 21 4. The Omissions of the Syriac Version 26 III. The Old Latin Versions 29 1. Both Old Latin Vei'sions Translated from the Greek 29 2. Both Versions of the Old Latin Exhibit a Mixed Text 30 3. The Current View that Old Latin h is Depen dent on Old Latin a is not Established by the Evidence Advanced 31 4. Old Latin b Independent of Old Latin a 36 5. Old Latin b is an Earlier Version than Old Latin a 41 6. Old Latin b Descended from a Greek Text which Antedates the Standard Text 44 7. The Latin Recension of Baruch according to the Codex Gothieus Legionensis 47 IV. The Greek Veesion 52 1. The Greek of Baruch Translated from the Hebrew 53 2. Only One Translator of the Hebrew into Greek 59 3. The Old Greek Text Originally a Part of Jeremiah 60 V. Composition and Date 63 1. Composition 63 2. Date 64 INTRODUCTION Under the title, The Principal Versions of Baruch, are included the Greek, Latin and Syriac versions. The Greek version has been preserved as a part of the Septuagint and follows Jeremiah in the chief manuscripts. Of the two Old Latin versions, one, 0. L. a, was incorporated in the Vulgate, and the other, O. L. h, is found in only a few manuscripts, distributed in Rome, Rheims and Paris. Manuscripts of the ordinary Syriac version are found in Paris and London, while the Syriac translation by Paul of Telia is preserved in a single Codex in Milan. These versions will be investigated in the reverse order to that of the chronological on account of the use made of the signs and marginal material found in the Syro-Hexaplar Codex. The following is a partial list of modern investigations of the Book of Baruch : Fritzsche, Handhuch zu den ApohrypJien, vol. 1 (1851); Kneucker, Das Buck Baruch (1879); Gifford, in Speaker's Commentary (1888) ; Marshall, in Hastings' Dictionary of the Bible (1898) ; Bevan, in Cheyne's Encyclopaedia Bihlica (1899) ; Rothstein, in Kautzsch, ApoTcryphen und Pseudepi- graphen des A. T. (1900) ; Toy, in, The Jewish Encyclopaedia (1902) ; Schiirer, Geschichte des jiidischen Volkes, 4th ed., vol. 3 (1909) ; Charles, in Encyclopaedia Britannica, 11th ed. (1910) ; Whitehouse, in Charles' Apocrypha and Pseudepigrapha of the 0. T. vol. 1 (1913). I. THE SYRO-HEXAPLAR VERSION. A literal Syriac translation of the old Greek version of the Old Testament was made from a hexaplar text by Paul of Telia in 616 A. D. The signs used by Origen together with many readings from Aquila, Symmachus, and Theodotion have been preserved in this translation. The poetical and prophetic books including Baruch are found in the Codex Syro-Hexaplaris which was published in fac-simile by A. Ceriani in 1874. The Syro-Hexaplar version of Baruch is more nearly identical with the Greek of cod. B than with that of any other codex. In a comparison of these two texts the Syriac shows the following differences, (1) additions: 1:20 ^'5^?l^'1 -r-^ = eK ytji Aiyv-iTTov ; 2 : 2 -» -s::^'" = d ; 2:18 ]¦-=' = TTViviJux ; 2 : 28 M = (TV ; 3:2 >a«94= ^M Uio-r^ lai^b ^S-^^ :^ OTL fleos eXeij/iOJv a km cAei/trov; 3:13 M^l^xpowv ; 4:8 T!^-? = Si, and under asterisks 3 : 1 la.^^» ,__k_i^i a-Sv-Lloii. 1^— 'c' = Tov ayayeiv e(l>'r}fM.i xa/ca fieyaXa. (3) omissions : 2:2 ev; 2:14 Trpoa-uiirov ; 2:18 avrov ; 3 :10 Ti ; 3 : 22 Se ; 4 : 33 fnoL, and (3) variations: 1:9 ^aJ) = avrovs, B avrov; 2:3 l-*l-=? = wwv, B vtou; 2:9 -? «^'»^s>=ois, B Ktti; 2:17 ^¦•Zlyj.ic? = Kara/3avT«, B reOvriKOTi's ; 2:17 l-o?] = SiKatcD/iara, B SiKotw/iui ; 2:18-? ¦•01=: 17, B 6 ; 2 : 39 |3 ^1 = ei ix-rj, B « /nijv ; 2:29 l-a-op^ = as fuiKpav, Beis^iKpav; 3 : 7 "^^ = etti, B aTro ; 4 : 8 ,;— ? = Se, B 8,; ; 4:13 ^? = 8e, B Kat ; 4 :14 -»iii2>? — =r'rJ° = TTtpte- Xei, B vepuXw. A comparison of the Syro-Hexaplar text with cod. A shows on the one hand that they agree in a few cases against B, but on the other that A differs much more from the Syro-Hexaplar text than B does. A and the Syro-Hexaplar text agree against B as follows : (1) inadditions: 1:18 ev; 1 : 30 ex yiys AiyvTrrov; 3:3d,- 3:11 xai; 3:19 rj/Jioiv ; 3 : 2 ort fleos eXcT/ZAwv et Kai eXeT/trov ; 3:13 )(povov ; 3:36 ot; 4:8 8e; 4:25 a-ov; (3) in omissions: 3:18 avrov; 3:10 Tt, and (3) in variations: 1 : 9 eto-rjyayev, B -qyayiv; 1:9 6 R. R. Harwell, avTovi, B avrov ; 3 : 3 vMv, B viov ; 3:29 w /MKpav, B as /AtKpav ; 3:7 cin, BaTTO; 3:23 owe, B otre; 4:8 St, B 8»?; 4:30 irapa- KaXet, B irapaKaXecret. Cod. A differs from the Syro-Hexaplar text as follows : (1) in additions: l:4aDTa)v,- l:5evxas; 3 : 4 t;ju,(i)v ; 2:13 87;; 2:14 rov ovopiaro's ; 2:17 Kat . . . . KvpLC .... trot ; 3:20 Xtywr ; 2 : 24 Kttt Ta ocrra twv ap^ovrtov rjp,o^-*', "it is finished". That these signs are not original in the Syriac, but have been trans lated from its Greek authority, is shown by their confusion in sev eral places. It is probable that a selection sometimes had two endings each marked by a >-*', so that the reader was at liberty to choose a longer or shorter lesson; for example, in Proverbs a ^t— ¦= is found at 10 : 27 followed by a ^^ at 11 : 2 and another at 11 : 7. But in several instances in Isaiah and Jeremiah there seems to have been" some rearrangement of text in the Greek authority; for example, a ¦^'r^ is found at Isaiah 32 : 14 while the next ^^ occurs at the close of 35:6; such a selection is by far too long and does not agree with the ordinary usage. Also in Jeremiah 31 : 9 the end of a selection is marked which has no beginning. Again it should be noted that a ^H" is not omitted when the pas sage selected for reading begins with the first verse of a book, for in the Wisdom of Solomon it is found at the beginning of the book. Now in Baruch each of the passages, 3 : 24 — 3 : 37, 4 : 1 — 4 : 4, 4 : 27 — 4 : 37, and 4 : 36 — 5 : 9, has its beginning marked by a -r° and its end by a'-^, but the end of a selection is also indicated in 1 : 5 which has no corresponding ^r-" . The natural inference to be drawn from this fact is that the book of Baruch in some of Origen's MSS. had a beginning which differed from that of the standard text, and that in recasting the text the ^t-° was lost which corresponded to the "-*- now found at 1:5. Although the latter had lost its meaning, it was still copied in the margins of later MSS. Finally, the note, "Not found in the Hebrew", occurs three times in the margin of the Syro-Hexaplar version. This note has been of uncertain interpretation because the subscription indi cates that the book was known only in Greek. The first occurrence of this note refers to the obelized words ^r,^-*-?^? ¦^^-^m "we who have sinned", on which J. J. Kneucker comments as follows : „Grundlos setzt der hexaplar. Syrer, verleitet dureh die Auslas- 8 R. R. Harwell, sung der Lucianischen Handschriften, die Worte t^^-'^? r-*^" unter Obel mit der Randbemerkung: hoc non positum m Hebraeo" {Das Buch Baruch, p. 223). Five cursives omit <«v rip.apropev, but Paul of Telia had before him a hexaplar text, if not the original Hexapla which was still existing, so that it is quite certain that these words bore the obelus in the Hexapla itself at the beginning of the seventh century A. D. Again, it is very probable that the original Hebrew of Baruch had dis appeared before Origen's time, or at least that he was not aware of its existence, for the Syro-Hexaplar version has a concluding note to the effect that the entire book bore the obelus, and with this the codex Chisianus agrees. Then, too, the case is further complicated by the fact that the obelized words in 1 : 17 are necessary to the sense and cannot he omitted, so that the pre sumption in favor of a clerical error is very strong. Following the suggestion of Nestle in Hastings' Bih. Did. vol. 4, p. 450 note, that the words, "Not found in the Hebrew", refer to the source, Dan. 9 : 8, and not to an original Hebrew text of Baruch, we find that a very slight error in placing the obelus in the Hexapla, or in a text copied from it, will explain the facts. For comparison the texts of Bar. 1:16, 17 and of Dan. 9 : 8 are placed in parallel columns thus : IJ'^tJ''? 1J*D'7D'? Kat rots /Jao-tXevo-tv rjp,v Kat rots ap^ouo-tv rjpoiv ; [IJ'XOJ/I IJ^JnD?!] Kat TOts t£p£vo-tv rjpu>v Kat rots irpo^rjrai'i -qpuiv ; *17 IJNtOn "1£J*N IJ'nDNvl Kat Tots irarpauiv r)pu>v (ov ¦qpaprop.f.v (vavri Kvpiov. The words in brackets are not in the Hebrew, so that it is hardly to be doubted that the Greek corresponding to these brack eted words should have borne the obelus instead of wv r]paprop.ev. The mistake might easily have been made in a series of phrases which begin and end with similar words, especially since the signs were written between the lines. A mistake in reading these signs was prevented in the O. T. by comparison with the Hebrew, but this was not possible for Baruch. This conclusion is corroborated by the somewhat similar instances in 2:3. Here the word avdpwirov which occurs twice bears an obelus in each case and has the same marginal note, "Not found in the Hebrew". It is not a question of a misplaced obelus, for in each instance the obelus has been placed upon the word that is lacking in the 0. T. source, which may have been one of the following passages : Dt. 28 : 53 ^1*70X1 Versions of Baruch. 9 7njm 7J3 "iti'n -[JDD ne, Lev. 26:26 lira Dn':'3N'i I'^DXn DD*nJ2 "ItJ^n DD*J3 • Baruch 2 : 3 agrees more closely with the second passage than with the first because "flesh" is repeated and the forms are plural, but in both cases the distribu tion of the subject by the use of JJ"N • • • • tJ"N is not employed as in the corresponding Greek of Baruch. Therefore the note, "Not found in the Hebrew", seems to refer to the source of 2 : 3 rather than to a Hebrew original of the book itself. Moreover, asterisks are found in 2: 1 with the clause, 1^='=' li-»-i= r:'^^ a-J-..iQl. "to bring upon us great evils", and this clause also occurs in the source, Dan. 9 : 12. The Syro-Hexaplar signs do not, therefore, give any basis for supposing a Hebrew text extant in Origen's time, nor any proof of a Hebrew original for Baruch. II. THE ORDINARY SYRIAC VERSION. The Syriac version of the canonical books of the Old Testa ment, called Peshitto, is based on the Hebrew; it is a generally faithful translation of a text practically equivalent to that of the Massoretes. According to Wright's judgment this version as a whole is a product of the second century A. D. {Hist, of Syriac Lit. p. 3.) The Syriac version of Baruch, however, exhibits several marked differences from that of the canonical books. It is based upon the Greek; and its final editor was guided in his work by the principle that the variants of his authorities must be preserved, with the result that the readings of different versions and recensions have been placed beside one another. Again, with regard to date the version in its present form is much later than the second century. During this period the Greek text of Baruch, on which the Syriac is based, was a part of Jeremiah, as the men who used the MSS. testify by their manner of quoting. But for the final editor of the Syriac translation of Baruch the separa tion of this book from Jeremiah was certainly in the distant past. To him the century that produced the Hexapla belonged to history, for the fusion of texts which it effected is perfect and all of its readings, both textual and marginal, are authoritative. Moreover that method of dealing with Biblical texts which might be called Lucianic, is not only established, but has reached its climax in the Syriac version of Baruch, in that all accessible variants are incorporated in the text. Ephraim Syrus (Opera Syr., 1743, III., 212) quotes Bar. 4:9, 19(?) from a Syriac version which appears to be the one from which ours is derived, but is without the conflations. In verse 9, 1 — 'i—^ and >Q^i^^ are wanting; in verse 19 the adjective i^^i--. The origin of these additions is discussed below; it is certain that they date from a time far later than that of Ephraim. Our 'Ordinary Syriac', then, is merely a late recension incorporating variant readings from many sources, as will be shown. The Syriac text has been published in the Polyglots of Paris and of London, both based on a manuscript in Paris. In his Versions of Baruch. 11 Libri V. T. apocryphi syriace, Lipsiae 1861, De Lagarde has based his text of Baruch chiefly on MS. 17105 of the British Museum. 1. The Syriac Version Translated from, the Greek. The Syriac version of Baruch abounds with additions and conflations, yet when these are removed it follows the standard Greek version. This expanded text of the Syriac is a serious objec tion to the hypothesis that the Hebrew original was among the authorities of the Syriac translator. It may be confidently asserted that there was never in existence a Hebrew original which contained the conflate readings of the Syriac, because the Greek version would have betrayed some knowledge of them if they had been in the Hebrew. The Syriac text, however, had a history before it assumed its present form. The first translator would never have so added to the text of his authority. Yet when these additions are removed, the remaining text is based on the Greek. It must be remembered also that the Greek of Baruch occupies a position between two cognate languages, the Hebrew and the Syriac, so that Semitic modes of thought and expression, though clothed in Greek, would be recognized by the Syriac translator and that words which had been inaccurately translated in the Greek would be corrected, especially if the Hebrew source which Baruch had used was accessible. Now it is well known that Baruch has drawn largely from Deut., 2 Isaiah, Jeremiah, Dan., and the Wisdom literature, and therefore his sources were prob ably familiar to the Syriac translator. Then, too, the Greek of Baruch contains a number of rare words which the Syriac translator would naturally interpret by comparing them with the sources if possible, so that the Hebrew of the sources, and not that of Baruch, will account for some of his renderings. The hypothesis, however, that the Syriac translator had before him, not only the Greek version, but also the Hebrew original of the first three chapters of Baruch, has recently been advocated by O. C. Whitehouse in an article on Baruch contributed to the first volume of The Apocrypha and Pseudepigrapha of the 0. T., edited by R. H. Charles, Oxford, 1913. This hypothesis is quite untenable since the Syriac text furnishes strong evidence that the Greek is the only authority of the Syriac translator. This is not 12 R. R. Harwell, intended to exclude the influence of versions that are also based on the Greek. The evidence which Whitehouse has advanced to establish his view is deduced from the following passages : 1 : 1 }.— i_A, "Shariali" The new name is not dependent on a Hebrew original as Whitehouse suggests, but it is due to a confusion of Daleth and Resh in Syriac, which was not infrequent. The loss of the initial Alaf is accounted for by SaSatoD, the reading of cod. A, which has also been fol lowed by both Latins. 1 : 1 ''^'-aiiii, "to Babylon". On this Whitehouse comments, "as though Baruch wrote the letter to Babylon. Heb. '7'2'2'2 one ^ dropped and '?DD was naturally interpreted as Accus. 'to Babel' ". The Syriac editor held the view that is stated in the Apocalypse of Baruch, 77 : 19, " And I [Baruch] wrote these two letters : one I sent by an eagle to the nine and a half tribes ; and the other I sent by men to those who were at Babylon". The text of this letter to the nine and a half tribes forms the conclusion of the Apocalypse of Baruch; the letter to the two and a half tribes is not given although it had been announced. The Syriac trans lator, or final editor, certainly regarded the Book of Baruch as the letter to the two and a half tribes. The evidence for this is the title and subscription of the letter to the nine and a half tribes together with the title and destination of the Book of Baruch; they are as follows: ir^*' ^=i-=? Ut^ Voiiii >aii^5cl ^ic jy-»,? "The letter of Baruch the scribe which he sent from Jerusalem to Babylon". The text of the letter to the nine and a half tribes is then given and is concluded with this subscription, >t°r^? li-»iej.B Uj-^^ ^'^'^*' li.^11? "End of the first letter of Baruch the scribe". Imme diately following is the Book of Baruch with the title, ^-Zizj U^^ "The second letter". This theory of the origin and relation of the two letters makes necessary the change from "Viias "in Babylon" to Vsa^. "to Babylon". 1:4 cTTt TTorapov 1iov8 The Syriac renders '=i? i'^'J V^io "and on the river Sur". This new name for an unknown river is advanced as evidence for the presence of a Hebrew original. The interchange of sibilants is no evidence at all, while the confusion of Daleth and Resh will account for the final letter of the Syriac. Versions of Baruch. 13 1:9 In this verse the Syriac has added U^=i "workmen", and has omitted Kat rov Xaov rrj^ yTjs. Whitehouse comments thus, "S varies considerably after 'princes', viz. 'and the officers and the workmen and the armies from Jerusalem'. Here 'the workmen' (rexwras = tJ*"inn) takes the place of 'the people of the land' in G and Syr.-Hex., and stands in closer accord with both Jeremiah 24 : 1 and the history of the year 597 B. C. (eomp. 2 Kings 24:14). We are in fact led to conclude that S here rather than G is based on the original Hebrew text". Does this mean that the Greek translator had before him the Hebrew text that is required by the Syriac, so that he translated tJ'"inn incorrectly by rov Xaov r-qs yirji while the Syriac translator rendered it correctly by l^^o] "workmen" ? The answer is self-evident, but in addition to this the suggested substitution must be rejected for the reason that rexvira^ is found in the margin of the Syro-Hexaplar, and the Syriac translator uses this Hexaplar material, not as a substitute for the text, but invariably as an addition to it. 1 : 19 eo-xeSta^o/icv is rendered in the Syriac by \'r^ "we rebelled". After stating the meanings of this word Whitehouse remarks, "Dan. 9:5, 11 enables us to restore the original". That it was Dan. 9, and not the Hebrew original of Baruch, which furnished the Syriac translator with his word, is evident from the following reasons: (1) In this verse the first word, ttTTo, has a variant reading, Ston aTro, which is found in seven cursives, namely, 22, 36, 48, 51, 62, 96, 231. The Syriac renders by t~*? ''*>^-J-^ "because from." This agree ment of the Syriac with these cursives is so frequent through out the book that it is difficult to escape the conviction that they have a common Greek authority. Again the rendering of aTreidovvre's by ^.^.w ¦ i ^^«^ |3 , "were not persuaded", as the Syro-Hexaplar regularly does, favors a Greek original for this word. (2) The influence of the O. T. sources of Baruch on the Syriac translator may be detected everywhere that it is possible to control him, but he reaches the climax in the seventh verse of the fifth chapter. The Greek varies from the source, Isaiah 40 : 4, by adding the phrase, Kat 6iva^ aevaovs, for which the Syriac translator substitutes from the source this sentence, i-*'"*^ le<^J l^or^o , "and the rough 14 R. R. Harwell, places shall become a plain". Here his leaning, and leaning heavily, on the O. T. source is evident in a passage that Whitehouse says is original in the Greek. (3) The Greek word under consideration occurs only here in the LXX., though the thought that is to be expressed is found repeated in Dan. 9 to which the Syriac translator could refer to find the exact meaning. 1 : 20 Kat eKoXXfjdrj £t9 ripa<; ra KaKa. For this the Syriac has li.A.j.s tt^«\s ji-^o "and there came upon us the evil". The relatively frequent occurrence of KoXXav in the LXX. makes quite certain the Hebrew original, tl^lil IJD pD"lJT) "But", says Whitehouse, "the Syriac rests on a variant nj^"in 1J3 N2fn which, has greater inherent probability since we have in Dan. 9 : 13 HND DUtn n^^in '?3 ". This is another case in which the Syriac translator leans on the 0. T. source instead of following his Greek. That the Greek translator had before him UNi which he mistrans lated by eKoXXrj6r] while the Syriac has it correctly rendered, is not at all probable since the same sentence occurs again in 3:4, and the Syriac translation ^a-^'o , "and there clung", shows that the Greek is not a mistaken rendering of the "Hebrew. 2:7 d (.XaX-qatv KvpLOi £ Tjpxx.'s. The Syriac renders, '^^-ic? VjLio \-'r^ ^-»1^ "inasmuch as the Lord has declared concerning us". Whitehouse comments, "This rendering of the relative in Hebrew HtJ'N] is not only more accurate, but brings with it better construction and sense. It is obviously not based on the Greek but on the Hebrew original". A com parison of this verse with its source, Dan. 9 : 13, shows that a relative pronoun should not be placed between raura and rjXSev, SO that the second d is plainly a dittograph from the last letter of ravra. It is correctly omitted by Q. In an effort to obtain a better construction six cursives have emended the first d to Stort, which has been followed by the Syriac with -? '^¦J^, "because". In the following list additional evidence is brought forward to show that the Syriac is wholly dependent on the Greek. 1:1 rov PipXiov = l?oi Ui~4? "this letter " ; 1 : 2 £v to. Katpo> = l^^}^ =01 "in that time"; 1:3 tovs Xoyous = •<;;-«^<'' V-ia-^-s "these words"; 1:3 r-qv ;8t/8Xov = V=r= ]-^<" "this book". Versions of Baruch. 15 The Syriac translator began his task by rendering the Greek article where possible with a demonstrative pronoun. He did not proceed very far before he discontinued this mode of translation, but these four cases are sufficient to show that he did not have a Hebrew original at the beginning of his work. 1 : 10 ayopacrare rov apyvpiov oXoKavroypara Kai irepi apaprias. Two kinds of sacrifices are here mentioned which are also found in reverse order in Lev. 5 : 7 H'^'J?'? inN") DNtSH'? IHH = £va Trept apapria^ Kat tva £t9 oXoKavrwpa of the LXX. It is scarcely conceivable that the Syriac translator could have so misunderstood the Hebrew original of "burnt-offering" and "sin-offering", when coordinated, as to omit the conjunc tion and render l"*-?^- '¦'^'^" liai^.-*- ]t-a.^ "burnt-offerings in behalf of sins". Only a misunderstanding of the Greek will account for this mistranslation. 1 : 14 Kat £1/ rjpepaK Kaipov. The Syriac renders, t-r^? livieo-^-so "and on the days of the Lord". This translation of Kaipov by J— 'r—^? "of the Lord", is satisfactorily explained as a misreading of Kaipov as Kvpiov. If the Syriac translator had had the Hebrew text before him at all, he would never have rendered by substituting "Lord" for "feast". According to his regular practice he would have added the word "Lord" to "feast". This fact is proved by his numerous additions, in which the Greek variants are incorporated in his text. 2 : 6 ripiv Se = •^— •? ^-^¦*? "but to us". Here the Syriac, instead of retaining its own idiom and that of the Hebrew IJ?! , as is done in 1 : 15, has followed the Greek. 2 : 18 £irt TO peyedo's is a mistranslation in the Greek which the Syriac has not improved with its oi-^a— ' "''^-^ "for its great ness". 2:35 n3T is mistranslated by airodToX-q as in Jeremiah 32:36, instead of by Oavaro^ as in Jeremiah 14:12; 6:7. The Greek is followed by the Syriac with V*"''*^ "in captivity". 3 : 3 'np . "dead of", is regarded by critics as a misreading of 'np "men of". The Syriac oi^-»i«? "-who are dead" agrees with the Greek. 3 : 8 o4)X-i)(TLv is found only here in the LXX. and it is rendered in the Syriac by Ua-o.*-- "debt", "condemnation". From 16 R. R. Harwell, the context, however, one expects a synonym of flOtJ' which Jeremiah has used so often in the same context. Hence, it is probable that the Greek has misread PlNJi'O "devasta tion" (comp. Job. 30:3; 38:27) as HNti'P "loan", "debt" (comp. Prov. 22:26). 3 : 14 /taKpo/3two-ts is found only here in the LXX., although the idea conveyed is regularly expressed by D*Q* 1"^^^ which the Peshitto renders by ]i^a-? li~,^CLi "length of days". There fore the influence of the Greek may be detected when paKpo^Liitai's is rendered by J-*^? li^ica^? 'f^^f^ "length of days of life". 3 : 21 ovSe := U-s] "nor". The same translation is also found in 3 : 22, 23, 27, 31, and 4 : 13. This indicates Greek authority as the Hebrew was probably H7'\ or VH") . 3 : 23 ot pvdoXoyoi occurs only here in the LXX. and the Syriac has followed the Greek closely with l^^^'O-*' ._»,i.i_»io5 ^,a!:^,<) "those who narrate myths". 3:24 £7rt/xijKr;s. Seven codices, 22, 36, 48, 51, 62, 96, 231, read evprjKrjs which the Syriac renders by ^-—o' t— •'' "abundantly long". 3 : 32 KTfjvojv Terpa-n-oSwv. The original was probably nOlD which the Peshitto constantly translates with ii-*^'=, "beasts", but here the Greek influence is unmistakable in 2>-i^']? Ii ''S''' "•"•¦^^-^l "beasts which have four feet". 4 : 1 aiSr?; Tj ^lySXos. In the Hebrew original this Greek would require IfiD Hf or ")3D DXt which must mean, "This [writing of Baruch] is the Book", so that the Greek is certainly a mistranslation. The avrr] should of course be avTTj from N'n , "She [Wisdom] is the Book". The Syriac translator was not influenced by the Hebrew original of this passage. 4 : 2 8ioSeva-ov. The Greek is reflected in o>-"ie|.s ^m "walk in its way". 4 : 7 dva-avTt^ Saipoviois. The effort of the Syriac to express this casual participle of the Greek is seen in bl-*'^ \0^-"'=?? ''^^ "by the fact that ye sacrificed to demons". 4:8 8£ Kat ^ -si t:^? "and also". The Hebrew possibilities DJ , flNI or DJl show no influence since the Greek is followed in Versions of Baruch. 17 word and position. Se is also rendered by .;:--? in 3 : 23, 34; 5:6. 4:9 yap = r— *— X "f or". The Hebrew is '^ and is usually pre- po.'iitive. The Syriac has followed the Greek in word and order. Exactly the same translation is repeated in 4:7, 10, 11, 15, 18, 19, 22, 23, 29, 33, 35; 5:4, 6, 7, 9. 4:25 piaKpodvprja-are. For this the Syriac has <=¦=!—=' Or^i"draw out your spirit". Since the Greek word is -used transitively, the Syriac inserts another verb to govern the object. 4 : 34 Most Greek codices have ayaXXiapa which has been corrupted to ayaXpa in A, 49, 90. The Syriac uses both, 15-?= b'Nv^, "idols and exultation". It is safe to say that this conflation is not due to a Hebrew original. Since in those passages which the Greek has mistranslated or inadequately translated the Hebrew, no aid has been given by the Syriac translator other than that which is accessible in familiar 0. T. sources, we judge that the evidence advanced establishes the thesis that the Syriac translator was wholly dependent on Greek authorities. 2. Additions to the Syriac Version Based on Greek Authority. The ordinary Syriac version has been characterized by Fritzsche thus : „Auch die Syrische schliesst sich ziemlich eng an ihren Grundtext an, nur wo sie sich nicht zurechtfinden kann, wird sie freier und umschreibt" {Die Apokryphen des A. T., p. 175). Kneucker elaborates this view and shows by means of many examples that the Syriac translator uses paraphrase not only where he is uncertain, but also expands the text in passages where there is no reason for doubt, and that he has a .special fondness for placing two synonymous expressions beside each other {op. cit. p. 168). A cursory examination of the Syriac version of Baruch dis covers many additions and doublets; a word or phrase of the Greek is rendered in the Syriac by two words or phrases, and in several instances by three. No explanation of this apparent verboseness has been attempted so far as I have been able to ascertain. Kneucker states the case, but offers no other explana tion than that the additions are due to the translator's exegesis and that the conflations are the result of his special fondness for 18 R. R. Harwell. synonymous expressions {op. cit. pp. 167 f.). This is equivalent to saying that they are arbitrary. That the Syriac translator, or editor, made additions to his text without MS. authority, is contrary to all that is known about the practice of translators of the Bible after its text had become relatively fixed during the third century A. D. Hence, this con jecture of Kneucker must be rejected. If now it is possible to demonstrate that a part of the additions to the Syriac text existed in the margins of the MSS. which were accessible to our translator, then the probability that he had authority for all his additions is strong. Fortunately in the margin of the Syro- Hexaplar text for Baruch there are found sixteen additions and variants, and one other occurs in the text under asterisks. Of these seventeen Hexaplar additions, the Syriac translator has used thirteen; of the remaining four only one could have been used by him without rejecting the text of the LXX. ; and the omission of this one may be due to homoeoteleuton. They are the following : 1 : 9 liicoPo " and the workmen ". 1:14 v^''^'^" "in behalf of us". This addition has its textual basis in r—'-*? lo"-^--, " our sins ", the marginal reading of the Syro-Hexaplar. 3:1 l^sJo? \li~t^i^ Uf^ t:-^^ ^•^1° "and the Lord brought upon us great evils". Except =, "and", and the subject, Up", "Lord", this sentence is found under asterisks in the Syro- Hexaplar text; according to a marginal note of Q it is a reading of Theodotion. O. L. a alone supports the addition of V-rio "Lord". 3:13 li^aio ^io "from a multitude". The Syro-Hexaplar reads l)-»-^ ^^, "from many". 3:15 o" . From the context this word means "art ". 3 : 26 l^Ja-i^ii. " for a devastation ". The Syro-Hexaplar has in its margin a synonym, viz., Wr^, "waste"; cod. 88 has epjj/xov under an asterisk. 2:29 ^oj^siJo "and they shall be diminished". This variant reading from Theodotion has been rendered in the Syro- Hexaplar margin by 1^^^ " to decrease ". 2:31 iJ^^oiB "understanding". This addition is also found in seven cursives. Versions of Baruch. 19 3:4 l-r* " Lord". This word has been substituted for o-ov of B; most cursives and Q have Kvpiov ; A omits entirely. 4:5 l^ilVj "of God". This reading is supported by seven cur sives ; B has p.ov. 4:6 ^n-n^n,\s^\ " to your enemies ". The Syriac is a combina tion of vjTEravrtous and IViVnSN "peoples", found in the Syro- Hexaplar margin; the suffix, \°-^- , is clearly implied, but has only the support of O. L. b. 4: 13 li^^a-o? " of truth ". This is a reading from Theodotion. 4: 18 Oi'*^!^, "upon you ". This addition is also supported by Q and nine cursives. To the above may be added one case from the margin of Q. 3:3 v^''^'^ >a*.»o tj] ) 1 Sn ¦¦! Sr> Vj.^ "because thou are compassion ate and pity us ". This sentence is also found in A and nine cursives; in Q, however, it is written in the margin. This last fact, together with its omission in B and the remaining cursives, is evidence that it is Hexaplaric material that has escaped the obelus. Therefore the Syriac version as it now exists has in its tex^ without Hexaplaric signs fourteen additions whose source is directly or indirectly the Hexapla as the marginal material of the Syro-Hexapla proves. It cannot be over-emphasized that this fact reveals to us the practice of the Syriac editor, namely, he incorporates in his text without signs the marginal additions and variants of his authorities. The Syriac text contains a number of shorter additions for which support can be found in the extant Greek codices. Their presence in. the Syriac also shows that it was the purpose of the translator, not to reject the variants of his authorities, but to save them. Those which are found in A Q, and in each case alsc in a part of the cursives, are: T^O -^'S^? *^'' "v^ "from the land of Egypt " ; 2 : 3 -? "which " ; 3:4 v-" us"; 2:14>aA, "name"; 3:171-1^ "Lord"; 3:36 -? "who"; 4:9 i"»-^"for"; 4:11 v^J) "them" (wanting in A); 4:25^ "your" (wanting in Q) ; 4 : 34 ^o8l^ " shall be ". A few additions have only cursive support. In the subjoined list the number of cursives that agree with the Syriac is indicated in the parentheses. 30 R. R. Harwell, 1:19 -j'M.io "because'" (7); 3: 8 ^ " thee " (8) ; 4 : 7 >='^i=^^? I^*^? "eternal God" (1); 4:9 -J "me" (2); 4:10 - "my" (1); 4:14 - "my" (2); 5:1 -^ "thy" (S-Hex. only). The additions which are found in the 0. L. and the Syriac prob ably had Greek authority ; the Latin version which agrees with the Syriac is indicated in parentheses. 3:1 l?o8i-^o "and Judah" (b) ; 3:9 ^r^iJ) "them" (a); 3:16 v- "us" (a); 3:11 -1^'^J " those descending " (a); 3:37a-3»i ' ' this is " (a) ; 4:1 vot"-^ v=°^^ "I will make an end"; (2) r=oi« "and I will cause to perish ", This conflation is found in O. L, a, " defectionem vestram faciam , . . . et auferam a vobis ". In Jeremiah 7: 34, the source of this verse, the LXX. renders Tl^tiTl by KaraXva-m and there are no variants, but in Jeremiah 36: 29 the same thought occurs, and the LXX. renders n'^UTF by ckXclij/ci while A Q 23 and some cur sives render by cKrpuj/a which the Syro-Hexaplar has translated in its margin by 1-==^, just the variant that the Syriac has added. 2:29 aTToa-rptij/a ets p,tKpav (1) Uojoi.-)^ ^a^ism^ " They shall turn to littleness"; (2) ^oj^lJo "and shall be diminished". The first sentence follows the Old Greek and the second is Theodotion's reading. 4:4 paKapioi icrptv la-parjX (1) ^J-" |J.sa-4 "Happy are we"; (2) '^-H*'*1 f^^<^4o. "and happy art thou Israel". The first part of this conflation agrees with the Greek and the second has been preserved in MS. corb. of O. L. b which reads, "Beatus es tu Israel". 4: 9 irevOoi ptya (1) l*' 11^1 "great mourning", according to most of the codices; (2) >aiii.:ii.» "(mourning) which is forever", according to a variant found in seven cursives. In this conflation the Syriac combines the two readings by means of a relative pronoun. 4:12 Oiov (1) i-r^ "Lord"; (2) U^l "God". All authorities read Oeov except O. L. b whose "Domini" sup ports the first part of this conflation. 4:13 7rat8tas (1) Uo>i._ic5 "of the discipline"; (2) 1i^-Jk-a_D? "of truth ", Versions of Baruch. 33 The second part is Theodotion's reading which has been com bined with the Old Greek reading by means of the Genitive. 4 : 34 TO ayaXXta/ta (1) V^-*^->^ " images ". This is a translation of ayaXpa, a corrupt reading found in A, 49, 90. (3) ij-*?' "exultation". This is the correct reading, which is well supported. The next list contains twenty-three conflate readings in which the second part is a more literal rendering of the Greek than the first. The most satisfactory explanation of this class of confla tions is that the first part of the conflation is due to the original translator, whose renderings are often free and sometimes inac curate. Then in the history of such a version corrections were made in the margins of the MSS. until the time when the final editor combined into one text both the original translation and the marginal material. 1 : 6 Kd6a €Ka.-Q^>^^, "praise"; (2) iM=^=, "and glory". 2:25 £v popcjiaux (!) l^i--^; (2) )^*J«i=o. The first is ambiguous, "sword" or "waste"; the second must be "and by the sword ". 3 : 4 Ta KaKa (1) iM^^, " curses " ; (3) lsJi-«.=o, "and evils ". 3:7 ahiKuxv (1) ]!^^^-^, "evils"; (2) V'=^°, "and iniquity". 3:15 CK rov; dTjO-avpovs avrris (1) ey) (1) -^^i, "was revealed"; (3) ->-^)=, "and was seen ". 4:3 Kat Ta (Tvp^tpovra (1) A ^^? ^r*, "whatever good thou hast"; (2) t^ w^o.a.o, "and thy advantages". 4:3 Tiyv8o^av (1) ^^h^a-s^-i., "thy praise"; (2) trO-l=, "and thy honor ". 4:13 StKatu/xara avrov (1) -oioJ^oos , "his commands"; (3) -oioil-jo, " and his judgments". 4:16 airj/yayov (1) '^^^ , "take captive"; (2) orS'o, "and lead away". 4:21 £K 8vrao-T£tas (1) U-'^^ -i-*l ^, "from the strong"; (3) j.i.Ji.ii^A' ^fA '^°, "and from the domineering rulers". 4:24 ptra SoItjs (1) Uop.«.r3. "with joy"; (2) li—oiu^^so , "and with praise ". 4 : 25 /AaKpoflv/nT/CT-aTE (1) r,a.a-oj oj.^, " draw out your spirit"; (2) o-f£i^sso ^ "and endure". 4 : 27 vTTo rov £7rayovTos (1) U,^ ^Q-^, " with the Lord " ; (2) om ^ ^.k^si ,^n-i»\s -i-il?, "by him who has brought these things upon you". 4:28 £irto-Tpa^Zaia*4lc oillli^A v^u)^ " according to all thy kindness"; (2) t^-*'=^ (Ji^:" fAo, "and according to all thy good will ", Versions of Baruch. 25 3:1 KaLTTvtvpa aK);8tQ)v (1) Ua-.jj^5 ^..oj, "a spirit that is in dis tress"; (3) l^ifi 'r^', " and grieved bodies ", 3:2 tXtrippovrja-iv (1) jisajOj "understanding"; (2) ii-i^ii-o, "and wis dom". 3:20, 28 e.irurrTjpri'i (1) jlaajoj^ "of understanding"; (2) )^ioiu.?o^ " and of wisdom". 3:21 TToppta eytvrjdrjo-av (1) v'="^'*-^^ aa.-5|, "they became far"; (2) o?!^", "and rebelled". 4:1 avrr; r) ;8t;8Xos (1) l^i^ Min, "this is the book"; (2) Ursojo, "and memorial (record)". 4 : 2 ETTto-TpEc^ov (1) -isz] , " tum " ; (3) -a^izjc, " and be inclined ". 4:9 Tr;v opy,;v (1) 1h°», "anger"; (2) li^io-o, "and wrath". 4:11 pera .... wevOos (1) U^l^, "with mourning"; (2) l^.i^&^o, " and with groanings ". 4:13 £ir£j8i;o-av (1) 0-='?, "they have gone"; (2) n m .i jZ] o], "or they obeyed ". 4:16 rrjv povrjv (1) i-^i t"*.^, "the one alone"; (2) Uy^n *Vi\o, " and the solitary one ". 4:19 epripos (1) %^i-^, "deserted"; (3) 1^-?^°, " and desolate ", 4:33 Xv7rr]6r)a-£raL (1) ^a-^ZZ, " gbe shall be grieved"; (2) lr=Zo o«^, "and she shall be troubled". 5 : 1 Tijs KaKa)o-£0)s crov (1) ^3lvA.tsjj "of thy evil"; (2) ¦'' * ¦¦?=, " and of thy suffering". 5:7 (Greek wanting) (1) ) rq*\^ "a plain"; (2) 1za-a*.^c, "and level place ". The final list contains three cases of conflation to the second degree, which are highly instructive ; they further reveal to us the working theory of the Syriac editor with reference to his task, namely, he must incorporate the variants of his approved authorities regardless of the duplication of clauses. 1 : 18 Kat 7jW€i6r)iTap.tv avr:^o, "and we seek mercy". In the source, Dan. 9:18, Theodotion translates 0**7*52 by pnrrovpev, and the LXX. by Seo/ae^, which is just the distinc tion that exists between this and the third part of the con flation. (3) I'— 3 — »_aZ ^- ,__»_io»o^ "and we throw our prayer". This is the literal translation with which the " jactamus " of 0,L, b agrees. 4:36 enopcvOrjaav oSovs TpaxEtas (1) 1-'^*^ °^l', "they went into captivity ", which is a very free rendering. (3) a_a_iioio \^t^^ i2!L.kjkj! li-i.io]^^ "They went in hard ways of rough place(8)". A more literal translation than the first. (3) IZiLaxii^ Izu.'JoVs oi^zloj "They were scattered in difficult ways". Although the verb is taken from the following sentence, yet the use of waw shows that the Syriac editor intended that it should be joined to the preceding sentence. About sixty conflate readings have been enumerated in the foregoing lists; for seven of these support has been found in other versions for both parts of the conflation. These seven cases are important because they show that the readings of Theodotion and the variants in the recensions of the Old Greek have pro duced conflation in the Syriac. Did the translator or editor arbitrarily add those words which no longer exist in extant codices.? or did he find them in the versions and recensions that he used ? Since it has been established that our editor incorporates in his text the marginal material of his authorities, it is very probable that he would also use the variant readings of the texts that were accessible to him. 4. The Omissions of the Syriac Version. The Syriac translator has omitted only a few words of the Greek text. This fully accords with his purpose to save every variant possible. A misunderstanding of the text in a few Versions of Baruch. 37 instances and the common clerical mistakes will explain the most of the omissions. They are the following : 1 : 8 apyvpa. The Syriac text probably was, r-^-^f J^Jus? 1— Jj~i , "vessels of silver which he made"; the first two words end with the same letter, and the second and third begin with ^? and ^? respectively, so that the copyist might easily overlook the word "silver". 1:9 Kat TOV Xaov tijs yrjs- All the versions except the Syriac have this phrase, so that it was certainly before our translator who may have omitted it because it is not in Jeremiah 24: 1, the source from which this verse has been taken. Another possibihty is that "the people of the land" were identified with "the poor of the land", who were not carried into captivity (2 Kings 25 : 12). 1 : 10 Kat. The translator misunderstood TTEpt apapruis which led to his rejection of the conjunction. 1 : 30 (OS 17 rjptpa avrr). This phrase was omitted because it was understood to refer to the infinitive clause, "to give us a land flovidng with milk and honey", whereas it should refer to the first sentence of the verse, namely, "The evils and the curse clung to us". 3:4 £Kei. In verses 13 and 29 this word occurs in the same con text and is translated; hence a clerical mistake probably exists here, a confusion of EKEt with the following Kat. 2 : 8 TOV irpoo-o)irov. The next two words are Kvptov tov, so that acci dental omission is probable. 3:4 Sr). The Arabic and Coptic also omit. In 2 : 13 the Syriac uses waw to render S-r], but waw is excluded here by the preceding vocative. 3 : 7 Kat. The Syriac probably used a text in which this word was wanting as it is in A Q. 3:10 Tt is a corrupt addition in B ; it is wanting in A Q. 3 : 10 Et. The difference in punctuation will account for the omission of this word. 3 : 18 oTi. Reusch has conjectured that the original was 'JSt which was misread as '0 ; the latter did not suit the context and was therefore dropped. 3 : 37 vtt' is wanting in A Q ; it is probable that it was not in the authorities of the Syriac translator. 28 R. R. Harwell, 4 : 2 KaTEvavTt. The Syriac subordinates c^cotos to Xap\piv by means of the Genitive; the two prepositions are thus brought together and are well represented by''*' — i3_ca_^^ "over against ". 4 : 13 avTov. The conflation produced in the Syriac by Theodo tion's reading has led to the rejection of this word. 4: 14 Kat. The corruption of \°'^lJ, "let them come", into ^^\, "ye", has made the conjunction superfluous. The Syriac has also omitted PafivXiovoi (1:9), viro r-qv a-Kuav (1: 13), -qpwv (2: 27 and 3:36), a^ro (4:16), Se Kat (5:8). III. THE OLD LATIN VERSIONS. There are extant two Latin versions of Baruch ; one has a longer form of text which agrees closely with the standard Greek, and the other has a shorter form of text which exhibits considerable independence by its omissions and variations, and especially by a few noteworthy additions that find no support in other versions. For convenience we shall follow Fritzsche and indicate the longer text by the letter a and the shorter by b. Jerome did not translate into Latin the books of Wisdom, Ecclesiasticus, Baruch, 1 and 2 Maccabees, but he adopted for these books the Old Latin translation which he found in use in the church, so that the version of Baruch denominated a was incorporated in the Vulgate and has shared its history. On the other hand the shorter text of b is extant in only a few MSS. ; of these Sabatier has used cod. Ecclesiae Remensis, n. 1 ; cod. S. Remigii pariter Remensis, n. 4; cod. S. Theodericii ad Remos, n. 1; duo codd. Paris. Monasterii S. Germani a Prates; et cod. Orat. B. 7, editum a J. M. Caro, Romae, 1688. Both a and b have been published in parallel columns by P. Sabatier in Bibliorum sacrorum latinae versiones antiquae, seu vetus Italica et caetera, Reims, 1743, '49, Paris, 1751. In regard to a mixed text, published by G. Hoberg as Die dlteste Lateinische Ueher setzung des Buches Baruch, see section 7 below. 1. Both Old Latin Versions Translated from the Greek. It is the generally accepted view that the translators of a and b did not have the Hebrew original among their authorities. This view is undoubtedly correct, for the influence of the Greek upon the Latin versions is evident throughout the book while no direct influence of the Hebrew can be found. The mistranslations of the Greek are followed by both a and h; for example, 1 : 10 Kat -n-epi apapTia<;, pro peccato ; 2:18 em ro peyeOoi, super magnitudine mali (omitted by &) ; 2:25 aTrouroXy] ; emissione ; 3:4 TE^i/TyKoraiv, a mortuorum, b def unctorum ; 3:8 oXripoawrjV rqv rrapa TOV Oeov croi epxoptv-qv, a et vide jucunditatem a Deo tibi venientem. Much of the evidence advanced below to prove that b is not dependent on a likewise shows that the Greek was the only authority of the Latin translators. 2. Both Versions of the Old Latin Exhibit a Mixed Text. There occur in a and b many of the additions, omissions and variations of the Greek as they appear either in the text of A or in that of B. Some of these are doubtless due to accidents in the transmission of the respective texts, but this is an inade quate explanation for the most of these differences. a and b agree with A in using the following words and clauses which are wanting in B: 1 : 20 ek yi^s Atyvirrov, 3: 1 roiv ayayeiv scj) rjpai KaKa pcyaXa, 3 : 3 d, 2 : 30 XEytuv, 3 : 2 OTt deoi ckerjpuov Et Kat eXejj- (Tov, 3 : 35 Kat, 4:8 8e, 4:9 yap, 4 : 34 Eo-rat, 5 : 1 o-ot. a and b agree with A in the omission of three words found in B, namely, 2 : 18 avrov, 3 : 10 Tt, 3 : 37 vn-'. a and b agree with A in the following variations : 1 : 1 SoSatov, B Ao-aStov ; 1:10 avrovs ; 1:11 ovTOJv, A avTOV ; 1 : 14 ¦qpepa, A rjpipafs ; 3 : 3 vtov, A vtwv ; 2:23 TToXewv, A ttoXews ; 3:39 piKpav, A paKpav ; 3:33 otTE, A ovte; 4: 13 eyvoio-av, A Ec^vXaW; 4: 16 povqv, A px>voyevT]V ; 4: 34 ayaXXta;tta, A ayaXpa ; 5:2 atwvtov, A aytov. Versions of Baruch. 31 Readings from Theodotion are found in both Latin versions. According to Q-marginal Theodotion is authority for the addition in 3:1, rov ayayeiv t r]pMS KaKa p.tyaXa, which both a and b have; in 4:13 both Latin versions omit TratSEtas, the reading of B, and use aXrjdeias which is from Theodotion ; in 1:95 uses /3oo-iXeo)s, Theodotion's reading, instead of /8oo-tX£vs, the reading of B. Finally, both Latin versions have readings which among Greek codices are found only in the so-called Lucianic cursives, especially 36, 48, 63, 331, for example, 3: 5 a 6 8 curs, tytv-qd-qptv, B eyev-qd-q a-av ; 3:8ab5 curs, ravra, B tovto ; 4 : 33 a 5 4 curs, -qpoiv, B vpoiv ; 5:6 a b 8 curs. Eto-a^Et, B Eto-ay£t ; 3 : 2 a 5 7 curs, add y-q ; and in 3:8 a b 10 curs, omit oTt. 3. The Current View that Old Latin b is Dependent on Old Latin a is not Established by the Evidence Advanced. A comparison of the Latin versions with each other shows that they have many renderings in common, some of which seem to point to a dependence of one version upon the other. The Latin of b exhibits also a better style than that of a, and though the language of both versions is post-classical, yet that of h con forms itself less to the Greek text than that of a does. Likewise h uses fewer ecclesiastical terms. Without giving reasons Fritzsche has expressed his judgment that b is dependent on a, and Kneucker has formulated the evidence in favor of this view. No English scholar appears to have examined the Latin versions. Marshall summarizes Kneucker's discussion of them in this sentence, "b is a recension of a, improving its Latinity, altering some of its readings to agree with B, and indulging in explana tory comments" {Hastings' Bib. Diet. vol. 1, p. 254). White- house says that it is clear that h follows a in a considerable number of details {op. cit. p. 579). Charles, however, expresses himself more definitely as follows, "The Latin one in the Vulgate belongs to a time prior to Jerome, and is tolerably literal. Another, somewhat later, ... is founded upon the preceding one and is less literal" {Encycl. Brit. 11 ed., vol. 3, p. 474). It is evident that Charles has adopted the current view for he does not undertake to prove that b is founded on a, yet an examination of the evidence that has been advanced by Kneucker to establish the dependence of & on a shows that it lacks cogency; for in spite of the fact that b seems to have improved upon a in a few 32 R. R. Harwell, eases, it remains true that a follows the standard Greek text much more closely than b does, and is therefore the superior version as might have been inferred from its incorporation in the Vulgate. This however does not prove that b was independent, yet if b had had before him the text of a and the Greek presup posed by a, the production of an inferior version would have been a step in the wrong direction, especially if a better Latin scholar were doing the work. Fritzsche has characterized a as „ziemlich wortlich", and b as „eine latinisirtere Ueberarbeitung der erstern mit Zuziehung eines griech. Textes und etwas freier" {Handhuch zu den Apokryphen, vol. 1, p. 175). Kneucker quotes this characterization with approval and adds that „sowol aiissere als innere Griinde . . . bestatigen das Urtheil Fritzsche's, welchem neuerlichst auch Schiirer beitrat" {op. cit. p. 158). The external evidence is of course the patristic citations; the majority of these according to Kneucker agree more closely with a than with h. If it could be shown that the early patristic quotations from Baruch were in closer agreement with a than with h, then no further argument would be needed. But the fact is, that because of its Christological bearing the most of these quotations are from the third chapter of Baruch, and it is just in this chapter that the two Latin versions are so nearly identical that it is impossible to determine on which the citation depends. The one exception, however, is the long quotation of 4 : 36-5 : 9 by Irenaeus, which is extant in Latin that is considerably nearer to the text of h than to that of a. Kneucker brings forward also the statement of Augustine that the Itala "est verborum tenacior cum perspicuitate sententiae", which applies better to a than to h. This inference is doubtless true, but it tells us nothing about the relative ages of the Latin versions of Baruch. Until new mate rial is brought to light the question of the relation of a to b must be determined by internal evidence, and here also Kneucker's proof in each case is only the selection of one of several possi bilities. It will be necessary therefore to examine with consider able detail the passages on which Kneucker bases his argument. It must constantly be borne in mind that this hypothesis of an improved and consequently dependent version of b postulates the presence of substantially the standard Greek text before b. Versions of Baruch. 33 „Fiir unmittelbare Abhangigkeit der Vetlat. b. von Vetlat. a. sprechen aber auch innere Indicien unwiderlegbarer Art". This "irrefutable" evidence is found on pp. 160 f. of Das Buch Baruch, and is as follows : 5 : 2 „Die Uehersetzung 'gloriae aeternae' der Vetlat. b. beruht auf Misverstand von 'honoris aetemi' in Vetlat. a." Neither version follows exactly the Greek r-q's Soi-q's rov atiovtov, in each case the adjective agrees with the noun regularly used in each version; 8o^a occurs eleven times in Baruch, a renders it eight times by "honor", and b nine times by "gloria", so that if both had had identically the same Greek text for this phrase, the respective translations would have been made without dependence or misunderstand ing. Moreover, Irenaeus agrees with b. 3 : 14 ,,'LongiturnitaSj virtus et vita' der Vetlat. b. lasst sich nur aus 'longiturnitas, vitae et victus' in Vetlat. a. erklaren". It is impossible to see why the "virtus" of b can only be explained from the "victus" of a. It is true that "virtus" is probably a copyist's mistake, but this may have occurred in the copying of b from an earlier MS. If b were dependent on a, it should read "vitae et virtus" and not "virtus et vita". 2 : 7 By homoeoteleuton b has omitted a clause which Kneucker explains „durch die Annahme, dass das Auge des Verf assers vom erstern 'haec quae' in Vetlat. a. zum andern 'haec quae' ebendaselbst abirrte". According to Sabatier a has "haec quia", not "haec quae". The possibilities here are at least three, namely, (1) that b was copying from a as Kneucker claims, (2) that b made the mistake in the Greek MS. which also has the relative pronoun repeated, and (3) that the mistake was made within the MSS. of h which must have been recopied one or more times before the extant MSS. came into existence. Hence, the first possibility seems hardly "irrefutable". 3:23 „Wenn Vetlat. b., abweichend vom Griechischen sagt; 'f abularum inquisitores ; prudentiae viam et sapientiam non recognoverunt', so lasst sich dieser Text wiederum nur aus der Uehersetzung der Vetlat. a. 'fabulatores et exquisitores prudentiae ["et intelligentiae" ist Zusatz] : viam autem 3 <^ary of^ fYALE DMNiTY SCHOOL] 34 B. R. Harwell. sapientiae nescierunt' unter Beriicksichtigung des Griechischen ableiten und erklaren, und diese Abhangigkeit von Vetlat. a. springt um so deutlicher in die Augen, wenn man beachtet, dass Vetlat. b. in der ersten Halfte des Verses selbst oDvEo-ts durch 'intellectus', hier dagegen mit Vetlat. a. durch 'prudentia' wiedergibt". If h has abridged the "fabulatores et exquisitores" of a, might not this have been done from the Greek text which a follows verbatim? Or might not b have had a slightly different Greek text from that of af If & is dependent on a in this phrase, how is its independence of a and of the Greek in this same sentence to be explained? In a as in the Greek the following "prudentiae" must be construed as a Genitive limiting "exquisitores", while in b "prudentiae" must be construed with "viam". Moreover, "especially clear" evidence of the dependence of b on a is detected in this verse because odvectis, which is repeated, is rendered in b by "intellectus . . . prudentia", and in a by "prudentia . . . prudentia." It is not unusual for h to render by two different words the same Greek word which has been repeated in the same or in the next sentence ; for example, 3:15 avr-qi . . . av-rq^, a ejus . . . ejus, h ejus . . . illius ; 3 : 32 avr-qv . . . avr-qv, a eam . . . eam, h illam . . . cam ; 4 : 18 Tpo^Evo-avra . . . EK^pEi/^aaav, a nutrivit . . . nutricem, h salvavit . . . nutrivit. 2 : 18 „Die grosste Beweiskraft hat jedoch die Stelle Kap. 2, 18, wo der Text der Vetlat. b. augenscheinlich aus Elementen von Vetlat. a. in der Weise zusammengesetzt und verkiirzt ist, dass einestheils die in den Worten 'anima, quae tristis est super magnitudine (mali) et incedit curva' liegende Schwierigkeit iibergangen, andererseits aus der 'anima quae tristis est et infirma' und aus der 'anima esuriens' kurzweg mehrere — zwei oder drei — 'animae tristes et infirmae et esurientes' hergestellt werden". The Greek text of this verse is partly corrupt as Kneucker has recognized and emended, yet a follows the Greek closely and removes the difficulty by adding "mali". h agrees neither with a nor with the Greek. It is quite probable that h used a dif ferent recension of the Greek text from that which a used, yet if the translator of b did abridge, is it not more reasonable to suppose that he abridged the Greek in which there was a real Versions of Baruch. 35 difficulty rather than a in which the difficulty had been removed? If both a and b exhibited the same emendation of a corrupt pas sage, then dependence might be inferred, but this is precisely what they do not do ; each one goes his own way. Instead of showing with the "greatest cogency" that h is dependent on a, this passage reveals the independence of b. 4 : 15 „Und ganz gleicher Weise scheint der Plural 'gentes de longinguo, improbas etc' in Vetlat. b. aus dem in Vetlat. a. per epizeuxin wiederholten 'gentem' entstanden zu sein, es miisste denn (sehr unwahrscheinlich ! ) der in Vetlat. a. folgende Pluralausdruck : 'qui non sunt reveriti senem neque puerorum miserii sunt' (in Vetlat. b. verkiirzt : 'quae seni et juveni misericordiam non praestiterunt') in Vetlat. b. ruckwarts wirkend 'gentes' erzeugt haben". Again a follows the Greek text closely from which b varies. Therefore it is inferred that h is dependent on a! No one knows why h has used the plural "gentes" instead of repeating "gentem", but that the reasoning of Kneucker on this passage applies with as much force to the Greek itself as it does to a, cannot be doubted. The context makes very improbable any influence of a upon b, so let it speak for itself : e-rrqyayev yap e-ir' avrovs eSvo's paKpodev e6vo<; k.t.X., a adduxit enim super illos gentem de longin- quo, gentem etc., h Qui suscitavit ad eos gentes de longinquo etc. The Greek text could be recovered from a by retranslation, but not from h. According to Kneucker the six passages that have been examined, together with a better Latin style, prove that b is dependent on a. It is evident that Kneucker's deductions agree with a precon ceived theory of the relation of the Latin versions to each other. For in every case there are several possibilities of which one has been chosen, often not the most probable one, because it agreed with the view that had been adopted. Furthermore, Kneucker claims that Vetlat. b. verrath ihre Abhangigkeit von Vetlat. a. und ihr jiingeres Alter durch ihr besseres Latein {op. cit. p. 159). The evidence to show that b uses better Latin than a, he has formulated in eleven propositions which may be summarized as follows: b as compared with a (1) prefers the relative construction, (2) substitutes dynamic relations for mechanical ones, (3) removes the enallage of number, (4) 36 R- R- Harwell, changes the person, (5) chooses more correct expressions and relations, (6) seeks to abbreviate the narrative as much as pos sible, (7) sometimes omits the second or third member of repeated expressions, (8) substitutes a pronoun for an omitted abstract noun, (9) leaves pronouns untranslated, (10) removes circum locutions as superfluous, and (11) presumably exaggerations are omitted {op. cit. pp. 154 f.). It will be shown in the discussion of the next topic that the translator of b varies both from the standard Greek and from a in double as many cases as he improves upon the renderings of a. Five of the above propositions relate to some form of omission. In all these cases b has omitted the Greek as well as a, only the conflations of the latter are an exception. 4. Old Latin b Independent of Old Latin a. Next to its independent additions, the most striking charac teristic of h as contrasted with a and the corresponding Greek, is its large number of omissions. A few of these are doubtless due to the accidents of transmission, but this explanation does not apply to the most of them, and it is certainly not convincing to argue that the translator of h had all of this omitted material before him in the version of a and in the Greek text which a. implies. There are 2433 words in a and 2226 in the corresponding text of h. These omissions are found in all parts of the book with a minimum difference between a and h of seven words in the third chapter and a maximum difference of 117 words in the second chapter. This characteristic distinction between the two versions may be illustrated by their respective uses of the Divine name. In a there are seventy-six occurrences of the word "Deus" or "Dominus Deus"; six of these have been omitted by h, in six other cases one part of the compound, "Dominus Deus", has been omitted, and in three cases a pronoun has been substituted. On the other hand h uses the Divine name in five instances which are not found in a; four of these, however, have the authority of B. Another feature of the shorter text of h that shows its inde pendence of a, is its treatment of temporal designations. The indefinite phrases, "as this day" and "unto this day", occur in a and in the Greek seven times, 1 : 13, 15, 19, 20; 2:6, 11, 26 > Versions of Baruch. 37 two of these 1 : 13, 20, are omitted by b. The Greek and a have two definite dates : the first is, "in anno quinto, in septimo mensis in tempore quo ceperunt Chaldaei Jerusalem et succenderunt eam igni" (1:2). According to b Baruch wrote "post annum quin- tum, et mensem septimum", which might have been at any time during his subsequent life. The second definite date occurs in 1 : 8 which states that Baruch vdth the vessels of the Temple returned to Judaea "decima die mensis Sivan". This definite date has been omitted by b, whose point of view is quite different from that of a and of the standard Greek. In the use of the Divine name and of temporal designations a has followed the Greek, but a is without Greek authority for its conflate readings which have also lengthened its text. These con flations are not numerous, but are very important, because the fact that the translator of h betrays no knowledge of them is strong evidence that he did not have the text of a before him. They are as follows : 1:18 Kat -qireid-qaaptv. (1) et non credidimus, (2) diffidentes, (3) non fuimus subjectibiles. The first is probably the original translation, the second represents a marginal reading, and the third is from -qBer-qaapev, a variant which is supported by b and the Syriac. 2:19 -qpei^ KarafSaXXopev rov cXeov. (1) nos fundimus preces, (2) et petimus misericordiam. In the source, Dan. 9 : 18, Theo dotion has pi-n-rovpev and the LXX. has Seopeda; the same distinction exists between the two parts of this Latin conflation. 2 : 21 (KXtvaTE) TOV wpov vpuiv. (1) humerum vestrum, (2) et cervicem vestram. Although in Jeremiah the Greek always uses Eto-ayEtv or ep^aXXeiv rov rpaxqXov and never KXtvEtv lo^iov, yet it is quite certain that Jeremiah 27 : 8, 9, 12, is the source of this verse. The first rendering is exact, yet because it did not agree with the source, it is probable that the read ing of Jeremiah was first noted on the margin and then it found its way into the text. 2:23 eKXeifeiv rroi-qa(o . . . tjxavqv. (1) defectionem vestram faciam, (2) et auferam a vobis vocem. The two Greek translations of 'flSCn may have been eKXeuj/eiv and eKrpifeiv, variants which are found in Jeremiah 36 : 29, and in the Syriac conflation of this passage. 38 R. R. Harwell, 3:23 T-qv eis aSov, a cum descendentibus in infemum, h cum his qui sunt apud inferos; 3:14 i,a-q, a victus, h vita; 3:21 o8ov, a facie, fe via; 4:1 a-n-oOavovvrai, a in mortem, b morientur ; 4 : 3 Ta cTvp(j>epovTa, a dignitatem, h utilitates; 4:6 -n-apopyiaai, a in ira ad iracundiam provocastis, h irritastis; 4:9 e-7reX6ovcrav, a venientem, b super- venientem ; 4:12 KaraXeKJiOeiay, a desolatam, b relicta sum ; -qp-qpwd-qv, a derelicta sum, b desolata sum ; 4 : 22 croirqpo^, a salutari. Versions of Baruch. 39 h salvatore ; 4 : 24 oicnrep yap wv, a sicut enim, h nam sicut nunc ; 4:31 r-Tj a-Tj TTTftjo-Et, a in tua ruina, b in casu tuo ; 4:32 8£tXatat, a punientur, b miserae ; 5:3 (rqv, a suum, b tuum ; 5:9 -qy-qa-erai, a adducet, b praeibit; 1:14 avayvft)o-£o-(9£, a legite, h legetis; 1:20 ¦q apa, a maledictiones, b maledicto ; 1 : 22 -n-oi-quai, a f acientes, b f acere ; 2 : 14 Se-qa-ewi, a orationes, h orationem ; 2 : 24 riKova-apev, a audierunt, b audivimus; 2:30 aKovo-ovo-t, a audiet, h audient; 3 : 7 ETTt Kap8tav, a in cordibus, b in corde ; 3:21 eyevq6-qaav, a facta est, b f acti sunt ; 3:27 eBwKev avrots, a invenerunt, h dedit illis ; 3 : 35 eKaXea-cv avTovs, a vocatae sunt, h vocavit eas ; 4:5 dapa-eire, a animaequior esto, b constantes estote ; 4 : 14 pv-qpovqai^, a prudentia, h sapientia ; paKpo^uaais, a longiturnitas vitae, h longiturnitas virtus ; 3 : 23 ot eK^-qrovvre'i, a qui exquirunt, b quaerunt ; 3:32 Kr-qvav rerpairoSmv, 40 R. R. Harwell, a pecu^dibus et quadrupedibus, b pecudibus ; 3 : 33 o arroareXXmy, a emittit, h jussit fieri; TropEVErat, a vadit, b ortum est; vitj^kovo-ev avrw, a obedit illi, h obaudivit ; 3:35 eXap\pav, a luxerunt, h lucem dederunt ; 4:3 8o|av, a gloriam, b gratiam ; aXXorpua, a alienae, b advenae; 4:6 8ta 8e to, a sed propter quod, b quoniam; 4:6, 8, 9, 12 ^Eov, a Deum, b Dominum ; 4:8 rpoevaavTa, a nutrivit, b salvavit ; 4:9 irapotKot, a confines, h civitates ; 4:12 e-n-ixaiperta poL, a gaudeat super me, h in me exultet; 4:13 ^eov, a Dei, b illius; 4:19 ^aSi^ere, a ambulate, b abite; 4:20 e^eSvaap-qv, a exui me, b exui; £V£8vo-a/iijv 8£, a indui autem me, b et indui aaKKov, a sacco, h cilicium ; 4:22 aur-qptav, a salutem, b salvatorem Kat -qXdev, a et venit, b quia veniet ; 4 : 29 eira^ei, a adducet, b dabit 4 : 37 p-qpari rov aytov, a in verbo sancti, h verbo Domini ; 5:1 Ev8vo-at, a indue te, b indue ; 5:6 EtcrayEt, a adducit, h reducit ; 5:7 /SaSia-r], a ambulet, h ingrediatur. 1 : 10 TTEpt apaprias, a pro peccato, b pro peccatis ; 1:15 Kat epEiTE, a et dicetis, b legerunt ; 2 : 17 eX-qp6-q ro irvevpa, a spiritus acceptus est, b spiritus accepti sunt; SiKaiuipa tm Kvpiw, a justifi- cationem Domino, b justificationes tibi ; 2:18 ij/vxq, a anima, h animae ; 2 : 19 8tKat(o/iaTa, a justitias, h justificationem ; 2 : 29 EV Tois edveaiv, a inter gentes, b gentem; 2 : 33 o8ov, a viam, h vias; 3 : 3 -qpeis a-KoXXvpevoi, a nos peribimus, b ne perdideris nos ; 3:5 aStKtojv, a iniquitatum, b iniquitatis ; 3 : 22 -rjKovaO-q, a est audita, h auditum est ; w(j>d-q, a visa est, b visum ; 4:5 Xao';, a populus, b populi ; 4:5 pvqpovevaov, a memorabilis, b memorabiles ; 4 : 13 SiKauapara, a justitias, b justitiam ; 4 : 14 EX^arojcrav, a veniant, b venite; 4:15 edvo's, a gentem, h gentes; 4:16 rjp-qpuiaav, a desolaverunt, b desolati sunt ; 4:22 Kat -qXOev, a et venit, b quia veniet; 4:25 oij/ei, a videbis, b videbitis; e-inp-qa-ri, a ascendes, b ascendetis ; o-e, a te, h vos ; 4:26 eiropevd-qaav, a ambulaverunt, h ambulastis ; -qpB-qaav, a ducti sunt, b ducti estis ; 4:36 epxopev-qv, a venientem, h superventura est.; 4:37 Epxovrat, a veniunt, b venient. 1:8 ev Tcu Xa/^Etv avTOV ra aKe-trq oikov Kvpiov Ta e^evex^evto ek tov vaov awoarpeipai el's y-qv lovSa, ft cum acciperet vasa templi Domini quae ablata fuerant de templo revocare in terram Juda, b et ut acciperent vasa domo Domini quae ablata fuerant de templo terrae Juda in Babylonia ; 1:13 Kat ovk a-n-earpeij/ev, a et non est aversus, b si f orsitan avertatur ; 3 : 3 ort od Kad-qpevo's, a quia tu sedes, h tu qui regnas ; 3 : 30 oto-Et avrqv xpvaiov ekXektov, ft attulit illam super Versions of Baruch. 41 aurum electum, b praetulit auro electo ; 4:10 eirqyayev avrots, a superduxit illis, b super omnes induxit ; 4:15 EmjyayEv yap et avrovi, a adduxit enim super illos, h qui suscita^'it ad eos ; 4:16 OVK ijaxvvd-qaav rrpeafivr-q'; ovSe 7rat8tov -qXe-qaav, a non Sunt reveriti senem neque puerorum miserti sunt, b seni et juveni misericordiam non praestiterunt ; 4 : 17 Tt Swar-q po-qO-qaai vpiv, a quid possum adjuvare vos?, b non possum juvare vos; 4:18 o yap E-n-ayaytov ra KaKa vpiv, ft qui enim adduxit super vos mala, b sed Dominus, qui vobis haec mala intulit; 4: 25 paKpoOvp-qaare rqv opy-qv, ft patienter sustinete iram, b patientes estote quia per iram ; 4:27 to-Tat yap v/xtov VTTO tov ETrayovTos p.v£ta, ft erit enim memoria vestra ab eo qui duxit vos, b erit enim illi pro vestra captivitate memoria ; 4:28 iaarrep yap eyevero -q Siavoia vpoiv eievre's ^-qr-qaai avrov, ft decies tantum iterum convertentes requiretis eum, b multipliciter conversi quaeritis. This list contains more than ninety cases in which a follows the Greek text closely and the rendering of h is less exact. Nega tively, this is evidence that points to the independence of b, because with a and the corresponding Greek before him, it is extremely improbable that the translator of h would have so often chosen the less accurate rendering which from the point of view of the standard Greek text is in many instances not a rendering, but an arbitrary substitution. 5. Old Latin h an Earlier Version than Old Latin a. The renderings in which the Latin versions agree do not in themselves show which is the earlier version, and even the few special agreements of a with b give no clue to show which is the borrower. Similar Greek authority will account for ordinary Latin renderings which are identical, so that the earlier date of h must be inferred from the more general characteristics of the two versions, to which may be added the evidence of Irenaeus. The point of view of a is that of the standard Greek text; only in the conflate readings has a added to this text. It cannot be questioned that a is at least later than the time when the book of Baruch had been definitely separated from the book of Jere miah; this had taken place by the time of Origen's Hexapla as may be inferred from the Syro-Hexaplar version. On the other hand h is based upon a Greek text in which Baruch has not yet 43 R. R. Harwell, been separated from Jeremiah, for it begins with Jer. 52 : 12, and its point of view is different from the later Greek text in the following instances : 1:13 ft, ad aures Jechoniae filii Joakim ; b, ad aures Sedechiae filii Joachim. This reading of h is contrary to his reading in 1 : 8, yet it is probable that Greek authority was found for each case. 1:8ft, cum acciperet vasa . . . revocare in terram Juda ; b, et ut acciperent vasa domus. The Greek and a represent Baruch as receiving the vessels of the Temple in order that he might return them to Jerusalem on the tenth of Sivan. According to b it is the priests and people in Jerusalem who receive the vessels and not Baruch in Babylon, while the verb "return" and the definite date are omitted. 1:15 ft, habitantibus in Jerusalem; h, qui migratus est ab Jerusalem. This is an entirely different point of view. 1 : 19 a, patres nostros ; b, nos. 2:4a, eos sub manu ; b, nos subditos. In these two cases the speaker in h identifies himself with the exiles, which is contrary to ft and the Greek. It may be permitted to add here the bearing of b upon a distinction which Marshall has pointed out in the confession 1 : 15-3 : 8, and in regard to which Charles says, "Thus it is the Exiles clearly who are speaking in 2:13, 14; 3 : 7, 8. On the other hand the speakers in the confession in 1 : 15-2 : 5 are clearly the remnant in Jerusalem." {Encycl. Brit. 11 ed. vol. 3, p. 453.) The incongruity of making those who are in exile write a con fession of sin for those who are still in their native land, is justified by the Greek of 1:15 and 2 : 4. This point of view, however, is a later one which the authority of b did not have. According to b the confession in 1 : 15 was not for "the men of Judaea and the inhabitants of Jerusalem", as the standard text has it, but for "omni Juda, qui migratus est ab Jerusalem," that is, for those in exile. Likewise the distinction of the standard text in 2:4, "And he gave them to be in subjection to all the kingdoms that are round about us . . . where the Lord scattered them," is not supported by h which has only the point of view of the exiles, "Et dedit nos subditos omnibus regnis, quae in circuitu nostra sunt, . . . quo dispersit nos Dominus." Versions of Baruch. 43 Another general characteristic of a as compared with h, is that the translation of the former is more mechanical than that of the latter. Wherever it is possible to compare successive stages in a version, or successive versions, with one another, it is found that the earlier renderings are freer and more idiomatic than the later. For example, the Greek of the Pentateuch is superior to the mechanical renderings found in Kings and Jeremiah. The Syriac versions of the Gospels also show the same difference; the earliest version, the Old Syriac, is free and idiomatic, the latest, the Peshitto, follows the Greek closely, while the Curetonian occupies an intermediate position in time and in character of translation. If the same general rule applies to the Latin versions of Baruch, then the date of h is certainly earlier than that of a. The early Church Fathers frequently quote the last three verses of the third chapter of Baruch, but these quotations give us no help in determining the relative ages of the Latin versions because they are almost identical in these verses. Irenaeus, how ever, has quoted the last eleven verses of the book. This quotation in its Latin form agrees more closely with b than with a. In the following lists only identical renderings are given. Those common to a and Irenaeus are : 4:37 collect! . . . usque ad occidentem . . . sancti, 5 : 2 justitiae, 5 : 4 Deo, 5:5 in excelso . . . collectos . . . usque ad occidentem, 5 : 6 portatos, 5 : 7 ambulet. In 5 : 1 Iren. and a have added an "et" which b omits. The cases in which h and Irenaeus agree are : 4:37 venient . . . emisisti, 5:2a Deo . . . impone . . ¦ . gloriae aeternae, 5 : 3 tuum, 5 : 4 vocabitur . . . aeternum, 5 : 5 soils ortu, 5 : 6 Deus . . . cum gloria, 5 : 8 odoris . . . praecepto, 5 : 9 praeibit . . . laetitia. Moreover, Irenaeus and b agree in omitting seven words that occur in a, namely, 4 : 37 enim, disperses, 5:1 te, et honore, 5 : 4 tibi, 5:8 et. These omissions are strong evidence that b antedates a, because these omissions are not affected by the Latin translation of Irenaeus and because both Iren. and b do not have the conflation of a that is produced by the addition of "et honore" (5:1). We therefore conclude that the concurrence of these general lines of evidence proves that b is the earlier version. 44 R. B. Harwell, 6. Old Latin h Descended from a Greek Text which Antedates the Standard Text. It is the generally accepted view that during the second and third centuries A. D. there were in existence many widely vary ing versions and recensions of parts of the Greek Old Testament, but not so clear is the working of those harmonizing influences which finally in the fourth century resulted in approximately a standard text. By 100 A. D. the limits of the Hebrew canon had been definitely determined along with a fixed mode of interpretation. As the Greek Bible differed considerably from the Hebrew Bible both in extent of the canon and in the contents of the text, many new Greek versions of the Hebrew text appeared. The consequent confusion was increased by the distribution of Christian scholar ship among several centers of equal ecclesiastical authority. This condition existed for more than two centuries during which the most successful effort to bring order out of the chaos of differ ing Greek versions was Origen's Hexapla. This put an end to the original form of the old Greek- text. Yet even the Hexapla would not have exercised so great a harmonistic influence if the different centers of ecclesiastical authority had continued to be equal. In the meantime, however, the importance of Rome as the seat of the Church had been increasing until it became dominant in the fourth century A. D. The influence of the Roman Church had been placed on the side of Origen so that by the end of this same century there had come into existence by a harmonizing process a Greek text which may be called the standard text, vdth which all the extant Greek codices substantially agree. Then as time passed the divergent versions which lacked the approval of the Church vanished, but not without leaving behind some traces, of which the 0. L. h of Baruch is an impor tant one. For this book b has a relation to the Old Greek similar to that which the Old Syriac of the Gospels has to the corre sponding Greek. It is asserted by Kneucker that the additions, omissions and variations of b are due to the exegesis of the translator {op. cit. p. 156). This is equivalent to saying that b is not a pure version, but is a mixture of translation and commentary. There is one, and only one, justification for such an assertion, and that is Versions of Baruch. 45 the assumption that the standard text of the fourth century A. D. was the authority which b used. This assumption, however, is groundless. That h has no conflations shows that its text has not had marginal corrections made from other versions, which later crept into the text; that h has a shorter text shows that it has not been conformed to the standard text; that b has many independent additions and frequently a different point of view shows that it has escaped the harmonizing process which Origen's work began and upon which the dominant Church set her seal. It is admitted that the translator of b was a better Latinist than that of a, and without evidence we are not justified in asserting that he was less faithful to his authorities than a was. A casual glance at the Syro-Hexaplar text of Jeremiah with its well filled margins, or at the critical apparatus for Ezekiel in Swete's Old Testament in Greek, should suggest caution in accusing a trans lator of arbitrariness in a book that shared the history of the prophetical literature in Greek- during the early Christian centuries. The evidence that the Greek text which h represents antedates the standard text may now be advanced. The extant Greek text of Baruch is an expanded form with two hundred words more than those which the Greek authority of h contained. A few of these additions which do not appear in h are : 1 : 3 Kat ev too-t TravTos rov Xoyov rmv epxop£VX-qaiv. b et maledictum in gentes. c et in maledictum et in debitum. 3 : 34 ot 8e aarepe6eiar]. h nemo in me exsultet quia vidua relicta sum. c nemo in me exsultet viduam et derelictam. The above list could be easily increased from any part of the book; and to further emphasize the fact that the author of c adhered rigidly to his authority, two cases of unusual interest are added: Versions of Barucli. 49 4:28 8£KaTrXao'tacraTE emarpa(f>evre'; ^-qr-qaai avrov. b multipliciter conversi quaeritis eum. c decuplabitis tantum conversi quaerere eum. 4:31, 32 8EtXatot, -at (Cod. A 87;Xatot, -ai). b miseri, -ae. c manifesti, -ae. (2) More than two hundred words of the Greek text, which are not represented in h, have been regularly supplied in c from the standard Greek text. The determining factor in this har monizing process was a literal translation of the Greek without the influence of a, as numerous instances show. For example : 1 : 14 EV -qpepa eopr-q<; Kai ev -qpepai'i Kaipov. b in die solemni. c in die solemni et in die temporis. ft in die solemni et in die opportune. 1 : 20 yaXa Kat peXi o)S -q -qpepa avr-q. h lacte et melle. c lac et mele sicut hie dies. ft lac et mel sicut hodierna die. 2 : 2 VTTOKaTO) TraVTOs rov ovpavov. h sub caelo. c sub omni caelo. ft sub caelo. 2 : 18 aXXa Tj ^vxq i] Xv-KOvpev-q etti ro peyedo'i. b sed animae tristes. c sed anima anxia in amplitudinC. a sed anima quae tristis est super magnitudine mali. 2 : 27 Kat eironqaa0-q, 3 : 1 -jravrOKparwp, 3 : 28 avr(av, 4 : 15 yap . . . paKpoOev, 4 : 25 Trapa tov ^eov . . . avTOu, 5 : 1 Bo$-qs. (6) Worthy of note are the misspelled words in the text of c, for example: 1:20; 3:5 eKoXX-qO-q, eserunt for haeserunt ; 1 : 22 oixopeOa, ibimus for ivimus ; 2 : 1 tovs StKao-avras, iudicabunt for iudicabant ; 2:16 evvo-qaov, anue for annue; 2:25 Tronjpots, multis for mails; 3:2 Et, est for es ; 3 : 6, 7 aiveaopev, laudavimus for laudabimus ; 3:36 XoyuxO-qaerai, extimabitur for aestimabitur ; 3:29, 30, 37; 4 : 1 avr-qv, eum for eam ; 4:15 Trat8tov, iubinem for iuvenem ; 4 : 17 Po-qO-qaai, adiubare for adiuvare ; 4:21 Trpos rov Oeov, ad eum for ad deum; 5:4 KX-qO-qaerai, vocavit for vocabit(ur) ; 5:5 Kat arqOi, et esta for et sta ; 5:9 -qy-qaerai, praeivit for praeibit. To these may be added one transliteration, "heremum", for eprjpov (2:26), and one case of corruption. In 5:6 c renders o)s Opovov ySao-tXEtas by sicut horum regem. This "horum" is an evident corruption of "thronum". The initial t was lost because of the final t in the preceding word ; then the impossible "hronum" was read as "horum", which in turn required the change of "regni" into "regem". Evidence was advanced in the discussion above to show that b represents a pre-Origenic Greek text, and that its production is incredible after the standard Greek text had been approved by the Church and the corresponding text of a had gained currency (see section 6 above). Even more incredible would be the pro duction of b with both the standard Greek text and the text of c before its author. On the other hand b with its distinct traits of an early translation, and c with the fundamental part of its text identical with h and with the clearest evidence of having been harmonized with the later Greek text, are conclusive proof that b antedates c in its present form. IV. THE GREEK VERSION. The book of Baruch is found in four uncial manuscripts, namely. A, B, Q, 23, and in twenty-two cursives. Its text according to cod. A was published by Holmes and Parsons in their edition of the Septuagint which contains the variant read ings of the remaining manuscripts except those of cod. B (Oxon. 1798-1827). Tischendorf also included Baruch in his edition of the Septuagint, which was first published in 1850. His text is a revised Sixtine; the latter, though based on cod. B, is an edited text. The most satisfactory edition, however, is that of H. B. Swete who began publishing The Old Testament in Greek accord ing to the Septuagint in 1887. The text is that of cod. B with the variant readings of the chief uncials printed below the text on each page. The book of Baruch follows that of Jeremiah in the third volume (Fourth Edition, Cambridge, 1912). The chief differences between the texts of A and B have been enumerated in a comparison of them with the Syro-Hexaplar version (see I. above). 1. The Greek of Baruch Translated from the Hebrew. The recent critics of Baruch are practically unanimous in their judgment that the first part of the book (1:1-3:8) has been trans lated from a Hebrew original. With regard to the second part, however, there is much difference of opinion as to whether it was composed in Greek or is a translation. Kneucker and Rothstein are convinced that the section, 3:9-5:9 also had a Hebrew original. Charles is inclined to this view (see note to White- house's article in Apocrypha of the 0. T., p. 573). Fritzsche vnth hesitation and Schiirer with positiveness decide in favor of a Greek original, while Marshall and Whitehouse grant a Semitic original for the first poem (3 : 9-4:4), but deny this for the second (4:5-5:9). The codices Syro-Hexaplaris and Chisianus have concluding notes to Baruch which state that the entire book was under the obelus, from which it is inferred that a Hebrew original was unknown to Origen. Also the note, "Not found in the Hebrew", Versions. of Baruch. 53 is probably not earlier than the Hexapla and is due to a com parison of the text with the sources in the Old Testament. Even the variants of Theodotion doubtless owe their existence to a revision of the Greek text. The external evidence, therefore, does not prove the existence of a Hebrew original. An examination of the text itself, however, shows that the entire book is written in translation-Greek, for it contains a suffi cient number of Hebraisms and mistranslations to prove that the Greek is a version. Important instances of words and phrases which require a Hebrew original for their explanation are the following: 1:10 TrEpt apapna^ is here coordinated with "burnt- offering " and is the Greek for DXCSPF , as its frequent occurrence in Lev. and Num. shows (comp. Lev. 5:7). 1:10 pawa. In Jer. 17:26 and 41:5 this word translates nilJO which the context requires here. The usual rendering is pavaa. 2 : 3 avOpioTrov is repeated in this verse and is a rendering of JT^N used in its dis tributive sense. The Hebrew construction in which "IJJ'J'J is followed by a pro noun, pronominal suffix, or adverb, to express the desired idea, appears in the Greek translation in several cases, namely, 2 : 3 ov . . . . ett' avTO), 3 : 17 o)v . . . . avru>v, 3 : 4, 13, 29 ; 3 : 8 ov . . . . EKEt. There occur in the text many Greek renderings which are mis translations and which can be satisfactorily explained only by a reference to the Hebrew original. Important cases are : 1:9 8£o-jao)Tas. In Jer. 34:1 and 39:2 this word is used to render "IJDD which has two different meanings, namely, "lock smith " and " dungeon". The second as equivalent to "prison ers " has been incorrectly chosen by the translator. 1:22, 2:21, 22, 24 epya^eaOat. In the second half of Jeremiah this word is used seven times with the meaning " serve" to render 13J7 (comp. 27:6, 9, 11, 13), instead of the usual rendering Bov- Xtveiv. The same mistranslation occurs four times in Baruch. 3 : 4 afiarov is properly an adjective and is regularly so used except here and in the second half of Jeremiah. In the latter it is frequently used to render HQiy' (comp. 44: 32; 48: 9). 3 : 12 ETrt Trao-tv Tots SiKaLwpaaiv aov. According to the source, Dan. 9 : 16 (LXX.) the original of this phrase is "jripiy "7^3 ,so that the e-n-i is a misreading of 3 as 3 . 3:33 e^uiOev. This is a mistranslation of myillD, "in the streets ", as is clearly shown by the occurrence of the same mis take six times in Jeremiah (comp. 33:10). 54 R. R. Harwell, %:%h airoaroX-q. In Jer. 33:36 the translator has rendered the Hebrew "151 . "pestilence", by this word which occurs in the same context. 2 : 29 Et prjv is doubtless corrected from Et p-q, the literal transla tion of N*? DN , used in oaths. The Syro-Hexaplar supports this view with P J, "if not". 3 : 4 reOvrjKormv is generally regarded as a misreading of ^piD > " men of ", as 'nO , "dead of ". 3 : 8 oX-qacv. It is quite probable that nNj^Q ¦ "devastation ", has been confused with TMiWf2 • "debt". T T ~ 3 : 10 £TraXat(i)d);s is certainly a mistranslation of n'72 which was used in late Hebrew as a synonym of JlJJ^ "afflict ". The Chroni cler has substituted ti72 in 1 Chr. 17 : 9 for ni^ of 2 Sam. 7 : 10. See also Dan. 7:25 (Aram.) and Theodotion's translation. 3 : 23 eK^-qrelv rqv avveaiv. This phrase in slightly different form is unnecessarily repeated. No one but a translator would have done this inelegant thing. 3 : 33 MEppav is unknown and probably owes its existence to a confusion of Daleth and Resh in |"IJ3 . In Gen. 37 : 38 the Mid- ianites are roving merchants. 3 : 23 b. 6 KaraaKevaaa^ . . . tverrX-qaev. The Hebrew obviously had the participle with the definite article, followed by an imperfect consecutive. "He who prepared . . . and filled . . . ". 4:1 Ets ^o)»;v. In the original Hebrew there was certainly the contrast of DTi? and rWU7 {a-TroOavowrai) . For the latter O.L. a has " in mortem". 4 : 5 pv-qpoavvov. It is probable that n^f , ' ' remember ", with the following "Yod" has been misread as the noun, f1"13t . 4 : 12 f . eK vopov Oeov . . . oSots evroXmv Oeov. The omission of the article before vojuov and evroXfuv points to the construct state in the original. 4:16 Kat aTTO TO)v Ovyartpiav ttjv povqv -qp-qpmaav. From Lev. 36 : 43, Ezek. 32 : 15, and Zech. 7 : 14 we learn that this Greek is probably from loatr rnWn hjDaV in Isa. 54:1 HODB' is used of a T woman. 4 : 36 £x^po)v This word probably points to a confusion of D'^Kf and DO'K • Versions of Baruch. 55 4:27 rov ETrayovros. This sort of ellipsis would be very natural in Hebrew poetry, but is not at all natural in this Greek. It is doubtless from npSH • 4: 28 8EKaTrX);o-ta(7aTE £Trto-Tpa<^£VT£s ^-qr-qaaL avrov. This sentence is ©neon which Whitehouse places special reliance, saying that "the phraseology is such as Hebrew could hardly employ" {op. cit. p. 573). It is at all events impossible Greek! The original Hebrew may have been ItTpD'? D'2B'' "W^ W^ ¦ Compare Jer. 17 : 11 and Rev. 3:18. 4: 31 f. 8EiXaiot, -at, -a would never have been written except by a translator. It probably represents DTlH which just suits here and is rendered by the corresponding verb in Jer. 30 : 10. 4:34 Ets TTEv^os. Taken in their context these words are unmis takably a translation of '^5^'? • 5 : 3 tj; Tratrjj. The origin of this feminine is plain enough from the Greek of Ex. 17:14, Deut. 25:19; 29:19, Job 41:3. The original was D'O^ Dnn b'2'^ ¦ 5:6 o)s Opovov jSao-tXEias is a mistranslation of n"li70 NDDD • The Hebrew will bear the translation, "As on a royal throne", while the Greek must mean "As a royal throne" (comp. Jdg. 33:36). The above cases of mistranslation are sufficient to show that the whole book of Baruch exists only in translation-Greek. In addition to this, however, the section, 3 : 9-5 : 9, furnishes strong evidence that it is derived from a Hebrew original by the fact that it consists of two poems. Several writers on Baruch refer to the second half of the book as poems though they claim at the same time that it is at least partly original in Greek. "The second part of the book (3 : 9-5 : 9), which differs greatly in form and tone from the first, consists of two poems. . . . That the first part of the book was originally written in Hebrew is prob able" (Toy in Jew. Encycl. vol. 2, p. 556). Marshall says of 4:5-5: 9, "This section is clearly divisible into four odes, each commencing with some form of Oappelv, and to these is appended a Ps. closely related to the 11th of the Ps. of Sol. . . . There is little reason to suspect that it ever existed except in Greek" {op. cit. vol. 1, p. 253). Whitehouse also states that, "In the rest of the book (3:9-5:9) we pass from the prose of the previous 56 R. R. Harwell, portion to poetry" {op. cit. p. 569). The last author carries out his view by publishing his translation of 3:9-5:9 in verses although he denies a Hebrew original for the second poem (4:5-5:9). What do these scholars mean by the use of the words, "poems", "odes"? Certainly they do not mean to assert that any part of Baruch is written in the form of Greek poetry which is based on a regular succession of long and short syllables, for that would be meaningless. Why then do they speak of poetry at all? It is because this section, though not metrical in the Greek, exists in a certain measured form which makes itself felt even in a Greek translation. There is an unbroken series of short sentences which betray their origin. And the fact that there are two poems, each in a different meter, is presumptive evidence that they were originally written in Hebrew. The first poem on the "Praise of Wisdom" (3 : 9-4 : 4) is didactic and was composed in three-beat measures, while the second poem on "Jerusalem's Lamentation over her Exiled Children and her Encouragement of Them" (4 : 5-5 : 9) was originally written in five-beat measures which was the regular poetical form used to express both joy and sorrow. A retranslation of the text into Hebrew has certain limitations, yet that it is possible to translate this Greek at all into Hebrew measures without doing violence to the thought expressed, is con vincing evidence that the Greek is not original. That any author would write out an original composition in Greek prose, even if the ideas were drawn from the Old Testament, so that it would fall into two Hebrew poems, each of which sustained an appro priate measure throughout, is scarcely within the realm of possibility. Although the entire Greek text of both poems has been translated into Hebrew measures, it has not been thought necessary to publish the complete retranslation. The first poem (3:9-4:4) was found to be practically free from later additions, and the second poem (4:5-5:9) has only a very few. As being representative ten verses from the beginning of each poem have been selected from our retranslation to show the true nature of the Greek. They are the following: {^iiri^'] 3:9 D'-^nr? nipn j;i^ Versions of Baruch. 57 m''3 mn'7 ^rrxn 3:9 T ¦ - - T • -: - •^Nitr' nr r\d7 3 : 10 ¦' T ; ¦ V T T D'm'2 nnx noeo -iin nil' d;; nati^m 3 : 11 •- : ¦ T : — : V noon "iipo nnrr 3 :i3 TIT I T : - T D'rt'^N T]-ina np'pr? dk 3: 13 D'i'iy'7 Di'^srn riDti''' T *" : T : T : - -^ nnnj 'N ni'^'Kno'? 3:i4 T ; " T • " Tin* ni^n'7 mnn ?« T : - *^ - T T : nilpQKVO'O 3:15 T I : T T .Tn'^yiK '?« X3 '01 T ¦.¦ : -.- T ' D'oyn ?'7jr'o n*N 3 : 16 D'PB^n liSV? D'pna'P 3:17 58 R. R. Harwell, 3njm fip^n n^'vi 3:i7 D'i^aa irican 13 im ' • T -: : T V ~: aim'? rrvp ]'k] D'J^fTj flPpn ^yin 3:18 "^Nntr' pn?r w -iN"i'ri "^n 4:5 Tpsj^'n*? n"? D'Iji'? D;nn3pi 4:6 Dnv"? Dr;i"|JpJ en*??* D5P»:i^pn iJ^! iD'rf7ii ii'y)) Dn^^'i D^rr^r D5's?'j; DmJ

^ 'N£2n ':';; 'sjn nnnn d'3"io 4: 13 -T : " T -: ^- • -: t ¦¦-: . - - ij;t ii'7 vcotDt^'Pi D'n'7N nnina no ^3 4:i3 :t t t : ¦ ¦ ..¦: - • t • D'n'7^{ niyp 'in-r? i^VC ^'^1 i3"n npni^n noio ninnN k'^i [|1*V'J.55?''] 4:14 D'711^ 'n*??* oh'b^, N^irr 7^¦^{ • 2. Only one Translator of the Hebrew into Greek. No convincing evidence has been advanced to prove that the Greek version of Baruch is due to two translators. This view has been frequently advocated because the prevailing opinion has been that the book is partly original in the Greek. Yet even Kneucker who is convincad that the whole book is a version strives to show that two different hands are to be found in the Greek. The evidence which he advances may be summarized as follows: (1) In the prose section Kat is frequently used to connect sentences, while in the poems 8£ is found or the conjunction is omitted. Also '3 is rendered by oTt and yap in the respective sections. (2) The Greek is more freely treated in the second part. (3) The peculiar Hellenistic words as well as differences in the mean ing and construction of words, point to a different translator for each half {op. cit. pp. 76-80). The differences, noted by Kneucker, exist, but they are not conclusive evidence, for he assumes that a translator using the flexible Greek language would render both prose and poetry in the same manner. More recently, however, Thackeray has proved that Jer. 29-51 and Bar. 1-2 have been translated by one hand, and he then adds to his discussion as self-evident the statement that, "The Greek of the latter part of Baruch is of an entirely different character, and is certainly by another hand" {Jour, of Theo. Stud. vol. 4, p. 264). 60 R. R. Harwell, The "entirely different character" of the Greek so far as it exists in the second part of Baruch is doubtless due to the "different character" of the Hebrew original. In the Old Testa ment a number of poems and fragments of verse are found embedded in a prose context. The difference between the Hebrew of these poems and that of the context is usually reflected in the Greek version. To illustrate, a short poem is found in the second chapter of Jonah. Though ideas relating to the sea, to prayer and to repentance are common to both the prose and poetry of this book, yet in the Greek sixty-nine per cent of the nouns, adjectives and verbs of the poem do not occur in the prose parts of the book. If the Hebrew original of Jonah had been lost, then the "entirely different character" of the Greek would have been sufficient evidence for a second hand. So also in Baruch, the change from prose to poetry in the original is a sufficient cause for a corresponding change in thte language of the translator into Greek. 3. The Old Greek Text Originally a Part of Jeremiah. In the discussion of the lectionary signs which are written on the margin of the Syro-Hexaplar text, it was found that a -*' for >ai:».^, "it is finished," occurred at Bar. 1:5 and that the corresponding '^i— *, "read," had been lost. On account of the regularity with which these signs are used throughout the Codex, this lost ^r" is evidence that there had been some recasting of the Greek text at the beginning of the book, for although the sign to indicate the beginning of the selection for public reading had been lost, the one to indicate the end was still copied in the margin of later manuscripts. Fortunately, it is not necessary to conjecture what may have originally preceded the opening of this book, for the manuscripts of O. L. b furnish positive evidence that the Greek authority on which they were based began with Jer. 51: 12. The last twenty- four verses of Jeremiah according to the Massoretic text have been abridged to nineteen in the Greek version. These are further reduced in the text of O. L. h to twelve verses which contain all the essential facts of the Hebrew original. To this addition of the Old Latin version is prefixed the title, "Incipit Liber Baruch". It may be reasonably inferred from this evidence that at an early Versions of Baruch. 61 date the books of Jeremiah and Baruch were written together in some manuscripts, and that when they were separated at least one division was made which differed from that of the later standard text. This view is favored by the nature of Jer. 52 which is clearly an appendix. Again, quotations from Baruch are ascribed by the early Church Fathers to Jeremiah. Both Irenaeus in Gaul and Clement of Alexandria wrote during the last quarter of the second century A. D. It is quite probable that the former used the Syro- Palestinian type of text, and the latter the Egyptian, yet both regard the second half of Baruch as a part of Jeremiah. The long quotation of Bar. 4 : 36-5 : 9 is introduced by Irenaeus with the words, "Significavit Jeremias propheta . . . dicens" {Against Heresies, Bk. 5, ch. 35). Clement of Alexandria also uses the phrase, 8ta rov lepepiov, to introduce each of his quotations from Bar. 3:13 and 4:4 {The Instructor, Bk. 1, ch. 10). About the middle of the third century Cyprian of Carthage prefixes to his quotation of Bar. 3 : 36-38 the words, "Item apud Jeremiam" {Testimonies against the Jews, Bk. 2, ch. 6). Therefore it is evident that no definite line of demarcation existed between the books of Jeremiah and Baruch in the manuscripts used by these Church Fathers. The book of Baruch was also included in the early lists of the canonical books in Greek. The seeming exception is in that of Melito whose list just antedates the time when Irenaeus and Clement of Alexandria were quoting Baruch as Jeremiah, so that it is very probable that Melito considered these two books as one. Only a half century later Origen likewise does not men tion Baruch in his canonical list, but the position of this book in the Hexapla proves that he included it with Jeremiah. From the time of Origen each successive list mentions Baruch with Jeremiah until the close of the fourth century when Jerome definitely separated the two books. Yet in the oldest Greek codices. A, B and Q, Baruch immediately follows Jeremiah. The lectionary sign copied from the Hexapla, the manuscripts of the Old Latin version h, the quotations of the early Church Fathers, and the patristic lists of canonical books, all corroborate one another in proving that Baruch was regarded as a part of Jeremiah during the second century A. D. 62 R. R. Harwell, The relation of the different versions of Baruch to each other and to the original Hebrew may be illustrated by the following diagram. [Hebrew] [Old Greek] Standard Greek [Theod. Recens.J Lat. b [Old Syr.] Syr-Hex. Lat. a Lat. c Ordinary Syriac V. COMPOSITION AND DATE. 1. Composition. The book of Baruch was originally composed in Hebrew and consists of two distinct parts; the first was written in prose (1:1-3:8), and the second in poetical form (3:9-5:9). The common theme which binds the two parts together is the destruc tion of Jerusalem and the exile of her children. The prose section is a mosaic of passages drawn chiefly from Deuteronomy, Jeremiah and Daniel. It consists of an introduction (1 : 1-14) and of a confession and prayer of the exiles (1:15-3:8). This first half of the book, written in prose, was probably composed by the man who edited the whole. Having before him this fine poetical composition (3:9-5:9), he ascribed it to Baruch and therefore wrote for it a suitable introduction. The section, 1 : 1-3 : 8, certaiidy appears to be the work of one hand ; observe how the request that prayer be offered in behalf of Nebuchadnezzar (1:11) prepares the way for the warning that national existence depended on submitting to him (2:21f.). The prose section as a whole has no Hterary merit other than that which is inherent in the borrowed passages. The second part of the book consists of two poems. The first exalts Wisdom, which Israel has forsaken (3:9-4:4). It is a unity and the thought advances gradually until it reaches a climax in the identification of Wisdom with the Law. The style is that of the Wisdom literature and compares favorably with that which is found in the twenty-eighth chapter of Job. The second poem is composed of four paragraphs, each of which begins with the exhortation, "Be of good courage". In the first three Jerusalem encourages her children, but in the last she is herself addressed in most inspiring language. This poem, though it has been marred in several places by early interpolations, exhibits an excellent s'tyle '"^th touches of the dramatic. The thought like wise rises to successive levels until it finally reaches the exalted strain of Second Isaiah. These two poems were probably composed by one author, for not only are they found together, but the second presupposes the 64 R. R. Harwell, first or its equivalent. The series of promises in the second poem find an adequate basis in the first. The reason why Israel has not been sold for destruction (4:6) and why God will restore him with joy and glory (5:9), is because the Law has been entrusted to him (3 : 37) and he knows the things which are pleasing to God (4:4). The cause of the captivity (3 : 12 ; 4 : 12) and the means of deliverance (4:2; 4:28) are the same in each poem. The entire contents of these poems favor the view that one poet, burdened with the misfortune of his people, chose Wisdom as an 'abstract theme and a personified Jerusalem as a concrete theme to express his thoughts on the exile and on the means of restoration with the purpose of giving encouragement. The literary ability of the author of the poetical section was far superior to that shown by the writer of the prose part. We repeat, therefore, that it seems more in harmony with the facts to assume that the poems were first in existence and that the author of the prose section gave them a definite setting by pre fixing a composition made up of excerpts from accessible Hterature. To give his work authority he ascribed it to Baruch in accordance with the literary fashion. 2. Date. The date of Baruch is uncertain. It cannot be fixed by a comparison of the fifth chapter of this book with the eleventh psalm of the Psalms of Solomon, for the ideas in both had been common literary stock since the time of Second Isaiah, nor is it convincing to assume that the relation ascribed to Nebuchadnezzar and Belshazzar in Dan. 5 and the prayer in Dan. 9 were new to the Jews at the beginning of the Maccabean age. Moreover, it is not safe to make prophetical and poetic language refer to definite events unless there is corroborating evidence. That every form of individual and. national suffering reached its climax in the destruction of Jerusalem by Titus, is evident, but it also requires no special proof to show that the Jews suffered and wailed over their hopeless political condition for more than six hundred years with the exception of a short period of independence under the Maccabees. Negatively, the book of Baruch shows no influence of the apocalyptic movement. The early view of destroying a city by Versions of Baruch. 65 fire from heaven and the Hebrew conception of a shadowy exist ence in Sheol ai-e affirmed, while the later view of a general over turning of the heathen as nations on the day of the Lord and a belief in the resurrection are wanting. The author assumes the point of view of the destruction of Jerusalem by the Chaldeans (1:2); for him the service of the Temple is being performed with regularity (1:14). Although the Temple is in bad condi tion (2:26), it is not necessary to suppose that it was entirely destroyed (comp. Pss. 74:7 and 79:1). The poem which exalts Wisdom belongs naturally to the period before the Christian era, and its composition after the destruction of Jerusalem by Titus is intrinsically improbable. Likewise the tone of hopeful ness and the promises of speedy victory in the second poem (4 : 24, 35) can hardly be reconciled with a date between 70 and 100 A. D. Moreover, a Hebrew original for the entire book is strong evi dence that the date of its composition precedes A. D. 70, for within two generations from that time the native church of Irenaeus in Asia Minor and that of Clement of Alexandria regarded the Greek version of Baruch as a part of Jeremiah. Again, the language of Jer. 29-51 in Greek shows an affinity with that of Bar. 1 :l-3 : 9 that is difficult to explain by imitation. Kneucker has observed that there is a "striking agreement" between the two {op. cit. p. 82), although he rejects the view that both had the same translator. The evidence to show that the Greek version of Jer. 29-51 and of Bar. 1-2 is due to one hand has been formulated by H. St.J. Thackeray who has pointed out an important distinction in the Greek of the two parts of Jere miah {Jour. Theo. Stud. vol. 4, pp. 26L266). The chief points of this evidence may be summarized as follows : Limited to Jer. and Bar. are the special meanings of Seap(or-q^s (1:9), epyat,eaOai (1 : 22), a^arov (2 : 4, 33), aTroo-roXi; (2 : 35), aTrotKto-/i,os (2 : 30, 32). Though rarely found elsewhere the following instances of agree ment are important: pawa (1: 10), Trpos ro p-q to introduce a nega tive purpose clause (1:19; 2:5), Trats instead of SovXos before "prophets" (1:20; 2:20, 24, 28), Ka^a (1:6; 2:2, 28), yEvos (2 : 15), e^oiOev (2 : 23), and the use of a participle with a pronoun dependent on rjpepa (2 : 28). The noun ^opp-qai's is found only in Bar. 2 : 29, and its cognate verb only in Jer. 31 : 35 ; 48 : 36 ; 1 Ch. 16:32. On the other hand Baruch has KXtvarE rov oi/tov (2:21) though Jeremiah uses Eto-ayayETE rov rpaxqXov (27:12). 66 R. R. Harwell. So many points of special agreement, therefore, are sufficient to prove that the latter part of Jeremiah and the prose section of Baruch were translated by one hand. Thackeray's evidence applies only to Bar. 1-2 to which he thinks we must give a much earlier date than that proposed by Kneucker and Schiirer. But if the entire book was composed in Hebrew, then, as the evidence fur nished above is sufficient to show, not only a part, but the whole must have been translated into Greek by a single hand sometime before the beginning of the Christian era. This view is cor roborated by the fact that the long quotation of Jer. 31 : 31-34, found in Heb. 8 : 8-12, reveals the use of our Greek text of the second part of Jeremiah. Hence, it is reasonable to conclude that Baruch formed a part of the Greek Bible before the rise of the New Testament. The date of the original composition would probably be not later than 100 B. C, and it is altogether reasonable to suppose that the poems were composed before the Maccabean age. YALE UNIVERSITY LIBRARY 3 9002 05032 9656 ^^vl < ' »• *"» * \ 's . ', ' \ ^ '• 'v '> '4 H *: -:;r-f 1 ¦ w A^^ k ' ^^* * ^P ^^ j^ ' *' ¥^ RC« M ' f*^ '¦^ . *!.«* > H^ .^' •;-t % ' A J^4 i .'«^* '^^M 't^ * It. * i?'*' '.'4^ t > .r* /? i ,.- . *>¦>. '*¦