§ 3. TEXT VERSIONS, DATE, ETC.

(a) Character of Translation. E, on account of its peculiar relationship to the O. T., cannot be studied textually apart from the versions based directly upon the MT (see more fully, Torrey 62–114). While the B of E-N is un-Greek, literal and mechanical, E is the very reverse of servile, and its language both elegant and idiomatic. The vocabulary is extensive, containing several words that occur nowhere else in ‘Septuagint’ Greek, or only in other books of the Apocrypha, notably 2 Macc. (see Moulton’s list, ZATW, 19. 232 seqq.). Semitic idioms are usually happily replaced by natural Grecisms. There is often a free treatment of the article, pronouns, and conjunctions; hypotaxis for the parataxis of MT; active verbs for passive. Condensation, paraphrase, and re-arrangement are frequent, and the translator has generally made the best of the original text, gliding over or concealing the difficulties. Sometimes he has misunderstood the original; but the rendering is carefully worded and thus presents an apparently plausible result (see e.g. 1:10–12, 38, 51). He manifests his intelligence when the skilful paronomasia ἄνεσιν καὶ ἄφεσιν (4:62), suggestive of a Greek composer, goes back none the less to a Semitic original (cf. Susanna, 54 seqq.), and the use of the name Sisinnes (6:3) in place of the MT Tattenai is typical of his care. E, it is clear, was made to be read, it is a version rather than a translation, and its value for the criticism of the MT must not blind us to its imperfections (on which see Bayer, 11 seqq.). Consequently, a mean must be sought between a promiscuous and haphazard use of E and a whole-hearted though indiscriminate reliance upon its readings and paraphrases. The attempt must invariably be made to distinguish between the underlying text and the features which (as in the Septuagint elsewhere) are due to the translator alone, and the difficulty of this task in certain crucial cases is vital for the disentanglement of the problems of E.1

(b) E and the MT. It is abundantly plain that E is not derived from the B (Theodotion) of the Canonical Books. Where there is agreement, the evidence points to accident or absence of intention, and is not strong enough to prove dependence (see the most recent study by Bayer, 156–61). In certain cases where they agree against the MT they sometimes are due to an easy misunderstanding, and sometimes point to a preferable reading; now and then the more literal version alone preserves an older text. It is highly significant that E is occasionally conflated, and presents simple doublets (e.g. 2:25, 6:29, 9:8, 46), or more elaborate combinations made with some little care (e.g. 5:50, 58, 72 seq., 6:5, 10, &c. [see Marq. 44–7]). This revision appears to have been made from the MT, and E 6:25 actually presents the incorrect ‘new’ (חדת) of the MT by the side of the correct ‘one’ (חד). Revision has also been made for the purpose of removing difficulties (so, probably v. 73b, in view of the date in 6:1), or of making identifications (Zerubbabel, 6:18, 27, 29). These adjustments, which are not found in Theodotion, seem to have been made first in the Greek version, and thus might appear to confirm the view that E is based upon an earlier Greek version (Ewald). The question of the underlying original, however, would still remain, and it is very important to notice that not only does E often presuppose a better text than the MT, but that some of the readings raise questions of literary structure and historical criticism. Consequently, E is not directly based either upon Theodotion’s literal translation or the extant MT; the marks of revision point rather to an attempt to adjust to the MT an earlier version which differed from it in some material respects, large (nos. 1, 2, 4, on p. 1 above) and small (e.g. 5:39 seq., 47, 6:28, 7:1, 9:38, 49).2

(c) MSS. and Versions. The Greek MSS. fall into two main classes, (1) Lucianic (MSS. 19, 108), and (2) B, A, &c. The former stand in a class by themselves, reveal many signs of correction and improvement in order to agree with the MT, and can be used only with great caution (see Torrey, 106 seqq.). The latter comprise two main subdivisions, B and A. B is distinctly the inferior, but shows fewer traces of correction. For a full grouping of all the MSS., see the elaborate discussion by Moulton, ZATW, 19. 211 seqq. א, it may be added, lacks E, but its subscription Εσδρας β (N 13:31) presupposes an ‘Esdras A.’

Two old Latin translations were printed by Sabatier (Bibl. Sacr. Lat. 3. 1041 seqq.), with a collation of MS. Sangermanensis—LC (Cod. Colbertinus; no. 3703), and a later which in a revised form was used as the Vulgate. A summary from a Lucca MS. was edited by Lagarde, Sept. Stud. 2. 16 seqq. (L Lag.). These differ from, and, on the whole, are purer than BL.

The Syriac Peshitta is without Chron., Ezra, and Neh. 1 EsdrA S is the Syro-Hexapla of Paul of Tella, printed in Walton’s Polyglot and by Lagarde (Lib. Vet. Test. Apocr. Syr., 1861). It is explicitly said to be from the Septuagint, and the same is stated at the head of a collection of excerpts in the old Syriac Catena, British Museum, Add. 12168 (see on 9:55). The variants of the latter and its selections from N are printed by Torrey, 5 seqq., and these selections, with a retranslation into Greek, collation, and complete introductory discussion by Gwynn (see p. 3 n. 3). S has many points of contact with BL, especially in 1:1–9, but on the whole a relationship with BB is more distinct.

The Ethiopic translation (ed. Dillmann, Vet. Test. Aeth., Vol. 5) represents the text of BB, S, &c., in contrast to BA, and, according to Torrey (101), ‘is a valuable witness to the Hexaplar text. It must have been made with unusual care from a comparatively trustworthy codex.’

The Arabic translation awaits study (PSBA, 24. 169); the Armenian is valueless (Volz, § 2).

(d) Josephus. The Jewish historian (first cent. a.d.), with his continuous history of the monarchy and post-exilic age, stands nearest (of extant compilations) to the chronicler in point of antiquity. He is a valuable exponent of the attempt to weave heterogeneous material into a readable and more or less consistent whole, and his greatest claim to attention lies in the evidence he furnishes for a comparative study of the traditions encircling the names and events of the period from Josiah to the Samaritan schism. Jos. is the earliest witness to E; the relationship is unmistakable as regards material and even language (Eichhorn, Einleit. Apokr. [1795], 347 seqq.; Treuenfels, Der Orient [1850–1]; H. Bloch, Quellen d. Fl. Jos. [1879], 69 seqq.). There are several points of agreement with BA as against BB (Thackeray, 762 b), and also with BL; Torrey (103) assigns the text an intermediate position. Unfortunately, Jos. is often extremely paraphrastic, and is therefore no safe guide for restoring the original of E. None the less, it is noteworthy that he is without the faults of E 1:29, 34 seq., he presupposes a text more complete and older than that in 6:18, 8:55, he uses a slightly different version of 3. (see Büchler, 64, 100), and, while obviously harmonizing in some places, elsewhere presents singular divergences or additions which do not appear to be arbitrary. In particular, his treatment of the stories of E and N is highly suggestive (see appendix to note on 9:55). Besides utilizing the canonical sources (Jer., Dan., Est.), he has had access (as in Est.) to other Jewish traditions (see on 7:15), and possessed some acquaintance with external history (see p. 11, and on 1:25). But although Jos. is not a direct witness to E’s text—and G. Hölscher has suggested that he made use of Alexander Polyhistor (Quellen d. Jos. [1904], 36, 43 seqq., 51)—he testifies to the authority of E’s history, and it is unnecessary to assume (Swete, Thackeray, Bayer, 140) that he used it simply because it was written in good Greek.

(e) Date and place. While Jos. is evidence for the earlier existence of E, it is not certain that it then had precisely its present form. As a translation the linguistic features suggest that it belongs to the time of the old Greek translation of Daniel, and was perhaps due to the same translator (Torrey, 84 seq.). The date of the original is bound up with that of Chron.-Ez.-Neh., and must be some time after 333 b.c. The Persian period was past, and its history had become obscure, the identity of Darius and Apame (4:29) was forgotten, and the points of contact with Dan. and Est. (not necessarily in their present form), would suggest the late Greek age. The problem also involves the question whether 3:1–5:6 is a secondary insertion or part of the original compilation, and this naturally affects the discussion of the home of the book (see pp. 29, 32). Although the section seems to some scholars to point to the influence of Alexandrian thought and philosophy (Lupton, André, Thackeray, Volz), to others it is Palestinian (Zunz), or not necessarily Alexandrian (Torrey). The identification of Apame speaks for Egypt or Antioch; the knowledge of the topography of Jerusalem (5:47, 9:38) is not that of the compiler or translator but of his source, and therefore cannot be claimed to support a Palestinian home. Egypt is suggested by the free irony in 3., 4., the unveiled women (4:18), the references to navigation (4:15, 23), and piracy (v. 27), and elsewhere by the use of Coelesyria (see 2:17). Thackeray (762 a) compares the ‘friends’ of the king (8:26; E B σύμβουλοι) with the ‘first friends’ who were third in scale of the courtiers at Alexandria, and with ἂν φαίνηται σοι (2:21, not in E) the phrase ἐὰν φαίνηται in Aristeas and frequently in Egyptian papyri. In so far as these data point to Egypt one may recall the interest in history-writing among the Hellenists Demetrios, Eupolemos, Artapanos, Alexander Polyhistor and others.

(f) English versions. It may be added that the old Geneva Bible, according to Lupton (6), is ‘in some respects closer to the Greek than that of 1611’. Various improvements to the A.V. are suggested by Ball in the Variorum Apocrypha, and even the R.V. is not such an advance as might have been anticipated. Note, for example, the archaic ‘Artaxerxes his letters’ (2:30), ‘cousin’ (3:7), ‘Jewry’ (5:7), and the gliding over of the obscurities of an imperfect B in 8:8, and especially in the concluding words, 9:55.