(1) Some scholars, as Ewald, Kneucker, Davidson, Rothstein and König, believe that the whole book was originally written in Hebrew; (2) Fritzsche, Hilgenfeld, Reuss, Gifford, Schürer, and Toy advocate a Hebrew original of i.-iii. 8 and a Greek original of the rest; (3) Marshall argues that i.-iii. 8 is translated from a Hebrew original, iii. 9-iv. 4 from an Aramaic, and the rest from the Greek; (4) and lastly, Bertholdt, Havernick and Nöldeke regard the Greek as the primitive text. The last view must be put aside as unworkable. For the third no convincing evidence has been adduced, nor does it seem likely that any can be. We have therefore to decide between the two remaining theories. In any case we can hardly err in admitting a Hebrew original of i.-iii. 8. For (1) we have such Hebraisms as οὗ ... ἐπ' αὐτῷ = אשר ... עליו (ii. 26); οὗ ... ἐκεῖ = אשר ... שׁם (ii. 4, 13, 29; iii. 8); ὧν ... τὸ πνεῦμα αὐτῶν = אשר ... רוחם (ii. 7). (2) We have meaningless expressions which are really mistranslations of the Hebrew. It is noteworthy that these mistranslations are for the most part found in Jeremiah - a fact which has rightly drawn scholars to the conclusion that we owe the LXX of Baruch i.-iii. 8, and of Jeremiah to the same translator. Thus in i. 9 we have δεσμώτης, "prisoner," where the text had מַסְגֵר and the Greek should have been rendered "locksmith." The same mistranslation is found in Jer. xxiv. 1, xxxvi. (xxix.) 2. Next in ii. 4 we have ἄβατον, "wilderness," where the text had שׁמה and the translation should have ἔκστασιν. The same misrendering is found several times in Jeremiah. Again ἐργάζεσθαι is used in i. 22, ii. 21, 22, 24 as a translation of עבד in the sense of "serving," where δουλεύειν ought to have been the rendering. So also in Jer. xxxiv. (xxvii.) 11, xxxvii. (xxx.) 8, etc. Again in πόλεων Ἰούδα καὶ ἔξωθεν Ἰερουσαλήμ the ἔξωθεν is a misrendering of בחוצות as in Jer. xi. 6, xl. (xxxiii.) 10, etc., where the translator should have given πλατειῶν. For βόμβησις (ii. 29) המון we should have πλῆθος. (3) Finally there are passages where by re-translation we discover that the translator either misread his text or had a corrupt text before him. Thus μάννα in i. 10 is a corrupt translation of מנחה as elsewhere in a dozen passages of the LXX. In iii. 4 τεθνηκότων = מֵתֵי - which the translator should have read as מְתֵי = ἀνθρώπων.
From the above instances, which could be multiplied, we have no hesitation in postulating a Hebrew original of i.-iii. 8.
As regards iii. 9-v. 9 the case is different. This section is free from such notable Hebraisms as we have just dealt with, and no convincing grounds have been advanced to prove that it is a translation from a Semitic original.