Song of Three Youths




The subject of this introduction is not really a ‘book’; and it is sometimes known as the ‘First Addition’ to the canonical Book of Daniel. It is an illustrative interpolation inserted in that book after 3:23; and is found there, forming an integral part of the book,1 in Theodotion, the LXX, Vulgate, and some other versions dependent on the LXX. It is absent from the Aramaic text.

It consists of four parts:

(a) Verses 1, 2. An introductory verse or verses connecting it with the narrative. This introduction occurs in two different forms in the LXX and in Theodotion.2

(b) Verses 3–22. A ‘Prayer’ ascribed to Azariah, one of the ‘Three Children’ who were thrown into the fiery furnace.

This ‘Prayer’ begins with praise to God (3) and an acknowledgement of His justice (4), especially in the judgement executed upon Israel (5–10). There follows a prayer for deliverance, for His Name’s sake (11); because of His promise to the Patriarchs (12, 13); because of their great sufferings and true repentance (14–19). The Prayer concludes with an appeal for deliverance, for the punishment of the enemies of Israel, and for the world-wide manifestation of the glory of God (20–22).

(c) Verses 23–27. A narrative as to the further heating of the furnace, the burning up of the Chaldeans round about, and the descent of the Angel of the Lord into the furnace to protect the ‘Three Children’.

(d) Verses 28–68. The Song of the Three Children with an introductory verse.

The Song is an ascription of praise to God, in which all His creatures, animate and inanimate, are called upon to glorify Him.

Verses 35–65 have the same refrain, ‘Sing His praise and highly exalt Him for ever,’ slightly varied in verse 52. These verses form the Benedicite.3


Just as this ‘Addition’ is not really a book, so originally, being merely a section of the Greek Daniel, it had no separate title. Thus Swete4 says, ‘In the Greek MSS. no break or separate title divides these Greek additions from the rest of the text.’5 But the Alex. MS. gives the Prayer and the Song under the titles ‘Prayer of Azarias’ and ‘Hymn of our Father’ as two of the fourteen hymns which it inserts as an appendix to the Psalter. Other MSS. head the Song ‘Hymn of the Three Children’. The Vulgate inserts after Dan. 3:23 the note Quae sequuntur in hebraeis voluminibus non reperi. Then follows our ‘Addition’ and then another note, Hucusque in Hebraeo non habetur; et quae posuimus de Theodotionis editione translata sunt.

Lagarde in his edition of the Syriac version of the Apocrypha gives the heading ‘Prayer of Hananiah and his companions’ from one MS., and from Walton (literally) ‘Prayer of the House of Hananiah’; the meaning of the latter phrase being the same as the former.6

In the A.V. and R.V. the whole ‘Addition’ is placed under the title ‘The Song of the Three Holy Children’, adding the note, ‘Which followeth in the third Chapter of Daniel after this place,—fell down bound into the midst of the burning fiery furnace.—Verse 23. That which followeth is not in the Hebrew, to wit, And they walked—unto these words, Then Nebuchadnezzar—verse 24.’

In the Prayer Book verses 35–65 form one of the canticles of the Morning Service, the opening words of the Latin version, Benedicite, Omnia Opera, being used as a title.

§ 3. THE MSS

The LXX version of Daniel was almost universally displaced at an early date by that of Theodotion, made in the first half of the second century a.d. The English versions are made from Theodotion.

The MSS. may be classified thus:

(a) MSS. of Theodotion’s Version1

A. Codex Alexandrinus contains the whole ‘Addition’ as part of Daniel and also the Prayer and Song as two of the Canticles.

B. Codex Vaticanus.

V. Codex Venetus.

Q. Codex Marchalianus, a complete copy of the Prophets preserved in the Vatican Library, written in Egypt not later than the sixth century. The margins supply copious extracts from the various Greek versions.2

Γ. Codex Cryptoferratensis, in the Basilian Monastery of Grotta Ferrata, a volume consisting partly of palimpsest leaves of an uncial MS. of the Prophets of the eighth or ninth century. This is available for Dan. 1:1–11a, 3:1–5a, 37b–52a, 7:1–8:19a, 9:15b, 26a, 12:4b–13.3

R. Psalterium Graeco-Latinum Veronense, ‘a bilingual Psalter of Western origin and attributed to the sixth century … the property of the Chapter of Verona,’ includes the Song as one of eight canticles supplementary to the Psalter.4

T. Psalterium Purpureum Turicense, a Western uncial, ascribed by Tischendorf to the seventh century, containing the Psalter, followed by canticles, including the Prayer and the Song. Verses 14–19 are wanting. It is in the municipal library of Zurich.5

Swete, pp. 165 ff., further enumerates a large number of cursive MSS. of, or including, Daniel.

(b) MS. of the LXX.

This version is only extant in the cursive MS. 87, the Codex Chisianus, in the library of the Chigi family at Rome. It contains Jer., Baruch, Lam., Ep. of Jeremiah, the LXX Daniel, Hippolytus on Daniel, Theodotion’s Daniel, Ezek., and Isaiah. It is usually assigned to the ninth century.6


(a) The Old Latin of Daniel is extant in various fragments and patristic quotations. These show that the version included our ‘Addition’. F. C. Burkitt’s investigations seem to point to the conclusion that before the time of Jerome there were current Latin versions of both the LXX Daniel and Theodotion’s Daniel.8

(b) The Vulgate of Daniel is made from Theodotion, and includes the additions.

(c) The Peshitta Syriac. Swete9 states that ‘From the first the Peshitta seems to have included the non-canonical books of the Alexandrian Bible except 1 Esdras and Tobit’. A. A. Bevan, however, writes that ‘The apocryphal pieces are found even in the oldest MSS. of the Peshiṭta, but seem not to have belonged to it in its original form’; this he infers from the statement of Polychronius, early fifth century, that the Song of the Three Children is not contained in the Hebrew and Syriac Bibles.10 The version is made from Theodotion, but differs considerably from both Theod. and LXX, probably through corruption and free handling.

(d) The Syro-hexaplaric Version is a literal version of the LXX column of Origen’s Hexapla, and supplements the testimony of Codex Chisianus as to the text of the LXX.

(e) A Syriac version of Daniel and other books of Jacob of Edessa, a.d. 704–5, exists in MSS. at London and Paris, but only specimens have been printed.11

(f) Egyptian Versions. (1) The Bohairic Coptic, based on Theodotion, but influenced by the LXX.12 (2) The Sahidic.

(g) The Ethiopic Version, based on Theodotion.

(h) The Arabic Version, based on Theodotion.

(i) The Armenian Version, from the text of Theodotion.13


This problem might seem to belong to the realm of pure scholarship; but it is involved in the controversy between the Protestant Churches and Rome as to the canonicity of the Apocrypha. Protestant divines have been inclined to regard original composition in Hebrew as one mark of canonicity, though they have never formulated any rigid doctrine to that effect. Dr. Barry, for instance, wrote of ‘the true Hebrew Canon of the Old Testament’.1 He probably only meant the canon current amongst Palestinian Jews, but the use of the term ‘Hebrew’ lays emphasis on the fact that the books of the Palestinian canon are extant in Hebrew or Aramaic; while the Apocrypha, when he wrote, were not extant in Hebrew.

Thus, according to Rothstein,2 most Protestant scholars since Eichhorn, including Fritzsche, Schürer, König, Cornill, and Strack, have decided for Greek as the original language; while Catholic scholars have held that the ‘Addition’ was written in Hebrew. But some Protestant scholars have also taken this view, e.g. Delitzsch, Zöckler, Bissell,3 Ball, Vatke, and Reuss. Further, Rothstein himself is inclined to accept a Hebrew original.

Bissell’s statement4 that ‘the majority of critics of all schools have always held to the opinion that this composition was originally written in the Hebrew or Aramaic language’ is too sweeping; but perhaps on the whole the balance of authority inclines that way. The theory of a Hebrew original is also favoured by J. T. Marshall:5 ‘The evidence for a Hebrew original is not irresistible, but probable’; Kamphausen6 states that it cannot be conclusively proved that the ‘Additions’ were written in Greek; J. E. H. Thomson7 argues for a Hebrew or Aramaic original; Swete8 writes, ‘The addition to Dan. 3:23 is clearly midrashic, and probably had a Semitic original.’

The present writer is clear that there was a Hebrew original probably for the bulk of our ‘Addition’, and certainly for the Prayer of Azariah. It must not, of course, be overlooked that various considerations, partly drawn from the study of the newly-discovered papyri, reduce the force of much evidence which would prima facie point to a Hebrew original. Idioms and words supposed to be Hebraisms or Aramaisms seem to belong to ordinary Hellenistic Greek, unless, indeed, the Jewish population of some districts gave a Semitic flavour to the local dialect. Moreover, it is always possible that if a Jewish author were more familiar with Aramaic and Hebrew than with Greek, or were soaked in the language of the LXX and had read nothing else in Greek, he might write original Greek as if he were translating from Hebrew. These considerations, accordingly, have been borne in mind, and due weight has been given to them; they lessen, but do not destroy, the force of the general arguments advanced, and there are specific items of evidence which are not affected by them. The conflicting views of various scholars show that there is not obviously an overwhelmingly strong case for either view.

Allowing for a very little editing or corruption of the text, there is not much that could not have been written in Hellenistic Greek, and nothing which could not…

Read more