The WYSS Website / Die WYSS Webseiten / Na Leathanaigh WYSS
Home / Heim / Abhaile



R. WYSS-Wolf

© R. WYSS 1979 / 2014

Complete as far as chapter 3.







Appendix: Line for Line Analysis of Daniel 224-278


Appendix: The Words for God in Daniel (Full List with Line References)



Appendix: The Proportion of On-Verses with Double Alliteration in Daniel


Appendix: Statistical Table Giving Detailed Information on the Distribution of the Verse Types



Appendices to Chapter III:

1) Index to the Scansion of Daniel

1) Notes on the Scansion of Daniel

3) The Scansion of Biblical Names in Daniel

4) Detailed Tables of the Distribution of the Verse Types in Daniel





The wish to write this study arose from a discussion of Daniel in Dr. Peter Lucas's course on the poems of the Junius manuscript which was held in the winter of 1978/1979 at University College Dublin. The reading of Hofer's article about the interpolation in the poem, and the discovery of a basic weakness in the latest argumentation by Farrell for the unity of the MS text were a good incentive for me to look into the question once again. In found that Farrell had overlooked a fundamental difference in the structure of the Vulgate text from that of the Old English Daniel, and I also thought that some of the results which Hofer and Steiner had obtained already in the year 1889 by the application of metrical and linguistic criteria could not be disregarded.

The outcome of my investigation is a vindication of Hofer and Steiner and refutation of Farrell's arguments.

I should like to thank Professor Bliss and Dr. Lucas for heartening me to undertake this study and for helping me readily with their advice whenever I have sought it.


in chronological order

Editions of Daniel


Bibliothek der angelsächsischen Poesie, 2. Band. Ed. Wülker, R.P. Leipzig 1894.

Facsimile 1929
The Caedmon Manuscript of the Anglo-Saxon Biblical Poetry, Junius XI in the Bodleian Library. Ed. Gollancz, Sir Israel. Oxford 1927.
Krapp 1931 / ASPR I
The Junius Manuscript, ASPR I (The Anglo-Saxon Poetic Records, Vl. 1). Ed. Krapp, George Philip. New York and London 1931.
Farrell 1974
Daniel and Azarias. Ed. Farrell, R.T. London 1974.
Works of Reference
Dn (405)
The Book of Daniel in Biblia Sacra iuxta Vulgatam versionem. Transl. Hieronymus, ed. Weber, R. Stuttgart 1969 
Sievers 1893
Sievers, Eduard. "Altgermanische Metrik" in Grundriss der germanischen Philologie, 2. Band. Ed. Paul, Hermann. Strassburg 1893.
Heusler 1925
Heusler, Andreas. Deutsche Versgeschichte mit Einschluss des altenglischen und altnordischen Stabreimverses. Berlin und Leipzig 1925.
Kuhn 1933
Kuhn, Hans. "Wortstellung und -betonung im Altgermanischen" in PBB LVII (1933), 1-109.
Pope, John C. The Rhythm of 'Beowulf'. New Haven 1942.
Sievers-Brunner 1965
Brunner, Karl. Altenglische Grammatik: nach der angelsächsischen Grammatik von Eduard Sievers. Tübingen 19653..
Bliss 1967
Bliss, Alan J. The Metre of  Beowulf. Oxford 1967 (revised ed.)

A Concordance to the Anglo-Saxon Poetic Records. Ed. Jessinger, J.B. and P.H. Smith. Ithaca and London 1978.

Studies of Daniel

Balg 1882
Balg, Hugo. Der Dichter Caedmon und seine Werke. Bonn 1882
Hofer 1889
Hofer, Oscar. "Über die Entstehung des angelsächsischen Gedichtes Daniel" in Anglia XII (1889), 138-204.
Steiner 1889
Steiner, Georg. Über die Interpolation im angelsächsischen Gedichte 'Daniel'. Leipzig 1889.
Farrell, H.T. "The Unity of Old English 'Daniel'" in Review of English Studies XVIII (1967), 117-135.
Farrell, H.T. "The Structure of Old English 'Daniel'" in Neuphilologische Mitteilungen LXIX (1967), 117-135.
Other Studies

Kluge 1884

Kluge, Friedrich. "Zur Geschichte des Reims" in PBB IX (1884), 128ff.

Schneider 1938

Schneider, Heinrich. Die altlateinischen biblischen Cantica. (Texte und Arbeiten herausgegeben durch die Erzabtei Beuron, 1. Abt. Heft 29-30.) Beuron / Hohenzollern 1938.

Hallberg 1962

Hallberg, Peter. Snorri Sturluson och 'Egils saga Skallagrimssonar'. Ett försök till språklig författarbestämning. (Studia Islandica 20). Reykjavík 1962.

Bliss 1971

Bliss, Alan J. "Single Half-Lines in Old English Poetry", Notes and Queries ns 18 (1971), 442-449.


Bliss, Alan J. "The Origin and Structure of the Old English Hypermetric Line", Notes and Queries ns 19 (1972), 242-248.

McTurk 1975

McTurk, R.W. Review of Um Fóstbrœδrasögu by Jónas Kristjánsson (Reykjavík 1972) in Medium Aevum XLIV (1975), 106-13.

Lucas 1977

Exodus. Ed. Lucas, Peter. London 1977.

Pope 1978

Pope, John C. "Palaeography and Poetry: some solved and unsolved problems of the Exeter Book" in Medieval Scribes, Manuscripts and Libraries: Essays Presented to N.R.Ker. Ed. Parkes, M.B. and A.G. Watson. London 1978.



In this thesis I seek to decide the question of interpolation in the Old English poem Daniel by reconsidering some of the arguments found in the works of Hofer, Steiner and Farrell.
The structure of this study is a consequence of the methodical approach. The handling of the source and the structure of Daniel, the words for God, and the quality of the metre are my criteria. Therefore my thesis falls into three independent chapters, whose order could easily be changed. I think, however, that it is best to begin with the investigation of the epic structure and to end with the rather lengthy discussion of the metre.
Since statistical evidence is also used to decide the question of interpolation, the problem of the significance of the results arises. Whenever Daniel A (lines 1-278 and 409-764) differs from Daniel B1 (lines 279-408) in any regard, it must be investigated whether the major sections of each part agree between themselves. If that is not the case, any difference between Daniel A and Daniel B may be fortuitous or owing to some other factor than different authorship. The two parts of Daniel have therefore been subdivided into the following sections:*

Occasional abbreviation
Extent (lines)
001 - 231
232 - 278
409 - 494
495 - 764
279 - 332
333 - 361
362 - 408

*Quotations from Daniel follow the text in ASPR I, edited by Krapp, unless otherwise stated. The editorial punctuation is omitted unless more than one line is quoted. The scansion of Daniel is also based on the text in ASPR I, unless otherwise stated in the "Notes on the Scansion of Daniel".



A comparison between the Old English Daniel and its Biblical source was undertaken by Steiner.1 He studied the handling of the Vulgate text by the English poet, especially in Daniel A, and found that the Song of the Three Children in Daniel B followed the order in the canticle version of the Espousing Psalter rather than the order in the apocryphal section of the Vulgate.2 Hofer also referred in passing to the Vespasian Psalter, but also found that the Song of the Three Children followed the Septuagint and not the Vulgate in the order of the verses, and that lines 399 - 408 of Daniel were taken from the canticle version in the Roman Breviary. The omission of LXX 28-3 (= Dn 405 3:52-56) in Daniel also has its parallel in the version given by the Roman Breviary, and Hofer concluded therefore that lines 362 - 408 in Daniel were a paraphrase of the Canticum trium puerorum as it is found in the Roman Breviary.3

What has not been given enough attention so far is the handling of the apocryphal matter by the English poet. Since Farrell denies that there is an interpolation in Daniel and holds that the poem follows the Bible quite closely,4 the relation of Daniel to its source or sources must be reconsidered.

It is possible that the poet used the Vetus Latina in one of its versions instead of or along with the Vulgate,5 as well as the canticle version of the Song of the three Children in the Roman Breviary. The use of different sources in different sections need not be a token of interpolation.

What is of the utmost importance here is the question how the anticipation of the outcome of the furnace episode that is found in Dn 3:22-24 is reflected in the Old English poem. Is it true that the repetition in lines 333 - 361 of part of the matter in lines 232 - 278 and the awkward anachronism of the Song of Azarias is about the same in Daniel as in the Vulgate version, as Farrell would have it?6

I shall first attempt to give a systematic survey that shows the characteristics of the treatment of the Biblical matter in Daniel A so as to establish the poet's manner of adapting and reworking his source. Examples will be mainly drawn from line 75 - 23l, but occasionally also from other sections.

I shall then undertake to hold lines 224 - 278 against the version in the Bible. Is it right to say that those lines are an anticipation of lines 333 - 361 that may be likened to the anticipation of Dn 3:46-50 in Dn 3:19-24? When we know how the poet handles Dn 1:1 - 3:18 in lines 75 - 223, we shall be better able to say what may be expected from him in his treatment of the furnace episode. Pope remarks that "lines 232-78 of Daniel include the entire narrative content of the apocryphal section ... alluding to but not quoting the lyrics".7 Is it likely that the poet of Daniel A skipped the lyrics; What are the contents and the structure of the furnace episode in lines 232 - 278?


The Poet's Adaptation and Reworking of Dn 1-3:18 in line 75-2318


1) The contents of Nebuchadnezzar's dream are anticipated in lines 111 - 115, although it is said in line 119 that the king does not know that he has dreamt.
2) The poet tells us that the king will be retaliated upon for his injustice (lines 186-7).


1) Daniel 1-75
The opening lines give us a survey of the history of the Israelites before the time of Daniel the prophet.

2) Daniel 85-87
The poet's auctorial comment may be seen in his remark that Nebuchadnezzar refused to bear in mind that he had to thank God for his gifts.

3) Daniel 014-107
The poem tells us about Nebuchadnezzar#s pride and fame, we learn that he used to lay down the law and was feared by the "children of men".

4) Daniel 166-167
The poet explains to us that nebuchadnezzar could not understand his own dream because of his sins ("for firenum").

5) Daniel 210-212
Nebuchadnezzar's anger at the Israelites' refusal to worship the idol comes better to the fore in Daniel than in the Vulgate, as Steiner remarks. Compare Daniel 210-212 with Dn 3:13!

6) Daniel 218b-224
The three young men tried hard to keep God's law, although they set their lives at stake by refusing to worship the idol.


1) Daniel's praise of God after Nebuchadnezzar's dream had been revealed to him (Dn 2:19-23).

2) Nebuchadnezzar's praise of the God of the Israelites after the furnace episode (Dn 3:95).

3) Nebuchadnezzar's praise of God after his return from exile (Dn 4:31).


1) Dn 1:10-17
Daniel asked Melzar, the supreme chamberlain, to give him and the three youths vegetables and water instead of the king's own food and wine. The wish was granted, and ten days later Daniel and the three youths were fairer in face and body than all the youths who had eaten of the king's dishes.
The passage is left out in Daniel because, in contrast to Dn1:6, Daniel himself is not introduced together with the three youths in line 91, but only in lines 145 ff., and because the incident does not help to develop the plot.

2) Dn 2:14-18 and 2:24-26
Daniel asked Arioch, the general, to introduce him to the king so that he could tell and interpret Nebuchadnezzar's dream.
Aroid is not mentioned in the Old English Daniel; Daniel himself goes to see the king to tell him his dream.

3) The scene where Daniel tells and explains Nebuchadnezzar's dream (Dn 2:27-45) is summed up in a single line:

sægde him wislice wereda gesceafte (line 160)

4) After hearing him tell and interpret his dream, Nebuchadnezzar kneels to Daniel and worships him, and admits that God of the Israelites is the God of gods. Daniel is given great presents by the king and made prince over all the land and all the wise men. At Daniel's request the three youths are appointed governors of the provinces of Babylon, and Daniel himself stays at the court of King Nebuchadnezzar (Dn 2:45-49).
The Old English poem only says that Daniel had great fame and renown among the learned men at Babylon. Nebuchadnezzar does not admit the superiority of the God of Israel, on the contrary, it is said that Nebuchadnezzar would not believe in God's might although Daniel tried to convert him:

Ne hwæðre þæt Daniel gedon mihte
þæt he wolde metodes mihte gelyfan (lines 168-169)

By this change the poet brings about, as Steiner remarks, a better link with the following episode than we find in the Vulgate.10


1) There is a long dialogue between Nebuchadnezzar and the wise men in Dn 1:3-11.
The element of direct speech is severely cut in the Old English poem (lines 130-144). The passage is not only shortened, but also reduced to an only sequence of question and answer.

2) The Vulgate passage Dn 3:8-18 is made up almost wholly of direct speech. In the corresponding passage of the Old English Daniel there is somewhat less speech, and it is all reported speech, with the one exception in line 208, where the poem switches over abruptly to direct speech in a relative clause.


1) Asphenez (Dn 1:3) is not mentioned in Daniel.

2) The new names that were imposed on Daniel and the three youths (Baltassar; Sidrach, Misach, Abednego (Dn 1:7)) are not found in Daniel.

3) Arioch introduced Daniel to the King. this character is dropped along with the whole passage 2:14-18 and 2:20-26.

4) The Chaldean men who accused the jews of refusing to worship the idol (Dn 3:8-12) are not mentioned in Daniel. We get the impression that Nebuchadnezzar himself noticed the Israelites' refusal and directly threatened to throw them into the fire if they did not change their minds (Daniel, lines 209ff.) The result is, of course, that the encounter between Nebuchadnezzar and the Israelites becomes much more direct and straightforward. The same may be said for the omission of Arioch (see above).

Omission of Dates and Numbers

1) Nebuchadnezzar told his officers to choose promising young Israelites for instruction in "cræft" (line 83), the Chaldean tongue, so that those young men could teach him their wisdom (lines 79-83), but the poem does not mention that the youths were taught for three years before they were brought before the King (Dn 1:5 and Dn 1:18).

2) The English poem does not mention that Daniel stayed with the King until the first year of King Cyrus' reign (Dn 1:21).

3) Nebuchadnezzar's first dream is not dated in Daniel whereas the Vulgate says that he had it in the second year of his reign (Dn 2:1).

4) The figures for the size of Nebuchadnezzar's idol that are found in Dn 3:1 are are left out in Daniel (lines 170-175). The figures may, however, have been found in the lost leaf.

5) Nebuchadnezzar orders the fire in the furnace to be made seven times hotter than usual (3:19). That numerical relation is not found in Daniel.


What then are the main characteristics of the poet's handling of the Bible text?

For one thing the poet helps us to understand his story. Therefore he gives us hints that are not found in the Bible. He uses anticipation to warn us that Nebuchadnezzar will not get away with his idol worship, and in sundry additions and amplifications he gives us information about the historical background of his story and tells us more about Nebuchadnezzar's character. He explains to us why this heathen king behaves as he does.11
The device of anticipation also helps the poet to stress the links between the different parts of the story. The poet makes for a smoother narrative by dropping the passages about Nebuchadnezzar's conversion and Daniel's appointment in Dn 2:45-49 and goes on to tell us about the idol instead. So the number of Nebuchadnezzar's relapses into heathendom is reduced by one.
the poet makes the thread of the story clearer by leaving out minor characters and episodes.12 These omissions also speed up the narrative, and the omission or reduction of direct speech has the same effect.
The omission of passages where God is praised may be explained in the same way, but another reason for these cuts is most likely that the poet shuns the lyrical elements because he wants to stress the epic character of his story.
The omission of dates and figures may again be owing to the poet's endeavour to stress the epic character of the poem: Daniel is not a chronicle!

The Structure of Daniel and the Relation of this Section to the Bible Text

The comparison of lines 224-278 with the Vulgate text is inadequate in Steiner's thesis and must here be done from scratch. The order of events in Daniel (as well as in Dn 3:19-24 and 3:46-50) is roughly as follows:

1) The King got angry (because the Israelites refused to worship the idol).
2) He ordered an oven to be heated and the three youths to be fettered.
3) The furnace was heated - it was an unusually big fire -
4) and the three youths were thrown in at the king's behest.
5) But God helped the faithful men:
    (5a) He sent an angel into the oven
    (5b) to word off the flames.
6) So the heat of the fire could do the youths no harm:
    (6a) They walked about in the midst of the flames,
    (6b) they were blithe
    (6c) and praised the Lord zealously.
7) The fire turned against the heathens instead, against those who had thrown the faithful youths into the furnace.
8) Nebuchadnezzar saw that a wonder had happened: the three youths were safe and hale, and an angel was in their company.

When we look at the Vulgate we find, as Farrell remarks,13 that the episode is told twice over. There is no full repetition, but the passage Dn 3:46-50 in the apocryphal part is an amplification of Dn 3:19-24. We must hasten to add that 3:24 is the first verse of the apocryphal section, a fact that Farrell overlooks in his argument. Farrell holds that there is a similar anticipation and amplification in the Old English Daniel.14 The question of interpolation can only be decided if we look into these things very closely and compare the "twice-told tale" in the Vulgate carefully with the Old English version of the story.

Dn 3:19-24 contains the following elements of my reconstruction of the whole episode: 1, 2, 3, 4, 6a, 6c, 7. It is understood, of course, that the fire does not harm the three youths (6) and that this is owing to God's help (5). But these things are not expressly said in Dn 3:19-24, whereas Dn 3:46-50 speaks of the angel that casts the flames out of the furnace (3:49), and of the youths that are unharmed (3:50).
Dn 3:46-50 contains the following elements of the reconstructed tale: 3a, 4, 7, 5a, 5b, 6, 6b ("neque contristavit" 3:50). That 6 and 7 appear in the inverted order is of no importance, since the two events may be said to happen at the same time. What happened before Azarias spoke his prayer (elements 1 and 2) is, of course, not repeated. We have the following additions in 3:46-50: 5a, 5b, 6b, so that the term "amplification" is applicable to this passage. It is not independent of 3:19-24 but gives further information.

What about the Old English Daniel?

Lines 232-278 can hardly be compared to Dn 3:19-24. We can say that lines 224-240 roughly correspond to Dn 3:19-24, and that lines 241-278 are an amplification of lines 224-240. But passage 224-240, unlike Dn 3:19-24, speaks with some emphasis of God's help and already mentions the angel that is found in the Bible only later on: in Dn 3, the "amplification". It is also expressly said in lines 233 and 239-240 that the fire could do the three young men no harm.
Lines 241-278, unlike Dn 3:46-50, refer back to the beginning of the episode: The heathen king was angry and ordered the tree youths to be burnt. Then come all the other elements of the reconstructed story, most of tem are repeated once or twice.
That part of the episode which corresponds to verse 3:22 and the apocryphal section of the Bible is framed by statements saying that it was God's help that saved the youths. lines 232-238, 240b and lines 277b-278. The first reference to God's intervention is, of course, also an anticipation of coming events.The second reference round off the episode and is something like the moral of the tale: the three youths were treated cruelly by the heathen king, but God did not let his faithful servants down.
The furnace episode is somewhat bewildering in its movement, and that may have been intentional. Things happened very quickly; the amazement of the three youths, and the fright and bewilderment of the King and his servants must have been colossal. The poet swings his camera from one spot to another, continuously changing the focus of his attention: from the angry King to his servants, on to the three youths, to God and his angel, back to the three youths, and so forth. It goes without saying that with all these repetitions and shifts of focus the actual order of events is not always followed.

Summing up we can say that the repetition and amplification found in the Vulgate is paralleled in lines 224-278 of Daniel. Of course, there are differences. The anticipation in the Vulgate comes before the Song of Azarias, and the amplification follows. The anticipation is rather slight, as Pope remarks.15 It is contained in the second half of Dn 3:22 and in 3:24. The thread of the narrative is taken up again in Dn 3:46, but the story does not begin from scratch.
In Daniel the anticipation is at once followed by the more elaborate account, which contains all the elements of the story, beginning at the very beginning. The episode is framed by emphasized references to God's help.

When we look at the whole episode we find that it is complete and need not be eked out by the account in the narrative bridge (333-461 = Daniel B2). The repetition in Daniel B is very awkward indeed after the very elaborate and complete episode in Daniel A. But even more awkward is the anachronistic insertion of the Song of Azarias after the full account of the furnace episode in the full account of the furnace episode in Daniel A. In the Vulgate the Song of Azarias follows upon the anticipation in Dn 3:22-24, but comes before the amplification.
If the Old English poet had treated the apocryphal matter in the same way as the Bible does, the Song of Azarias (lines 279ff.) would have to come after verse 234a, or else after verse 234a and lines 239-40, like this:

Gearo wæs se him geoce gefremede; þeah he hie swa grome nydde
in fæðm fyres lige, hwæðere heora feorh generede
mihtig metodes weard.
Ne mihte þeah heora wlite gewemman
wylm þæs wæfran liges, þa hie se waldend nerede.
Þa Azarias ingeþancum
hleoðrade halig þurh hatne lig,
... (lines 232-234a, 239-240, 279ff.)

In contrast to his source the English poet uses direct speech very sparingly. There are a few passages in the Vulgate where Daniel or Nebuchadnezzar utter words in praise of God, but, as we have seen, they are not rendered in Daniel. The English poet stresses the epic element in his paraphrase. So he may have felt free to skip the Song of Azarias and the Song of the Three Children, all the more so because Jerome himself warns us that they are apocryphal.
A further point must be made. The account of the furnace episode in Daniel A already mentions Nebuchadnezzar as a witness to God's intervention in lines 254b-255a and 268-9. This element is not found at all in Dn 3:46-50, but only in Dn 3:91-2.

If we assume that Daniel B belongs to the original poem we must take into the bargain that the poet has displaced this element quite considerably from its position in the source. But it is more likely that the poet wanted to prepare us for Nebuchadnezzar's reaction to the wonder in lines 409ff. If we skip Daniel B, this works out very well: lines 268-269 are then not very far from lines 409ff. But if we claim that Daniel B is no interpolation, then the element of anticipation in the references to Nebuchadnezzar in lines 254b-255 and 268-269 is almost lost.


On the grounds of my comparison of Daniel with the Bible text and on the grounds of my analysis of the structure of lines 224-278 I deem it unlikely that Daniel B is part of the original poem. If we take Daniel B to be an interpolation and remove it from the poem, we get a smooth narrative that is in keeping with the epic character of lines 75-231.



The following table analyses Daniel 224-278 line for line and gives the correspondences in the vulgate. The elements are numbered as in my reconstruction of the episode.16

Vulgate Dn

OE Daniel

1)  The King got angry.
2)  He ordered the furnace to be heated
     and had the three youths fettered.
3)  The oven was heated till the fire was very hot.
4)  The King had the three youths shoved into the blazing flames.
5)  God helped the three youths:
5a) He sent an angel into the oven to shield them.
6)  The heat of the fire could do no harm to the three youths
5)  because the wielder saved them.
1)  The heathen King was angry
20 (23)
4)  and ordered the three youths to be burnt.
22a, (46-)47
3)  The oven was heated, the fire was very big,
     the iron was thoroughly heated.
     the servants threw firewood into the flames
     as they had been told.
(46-)47, (22a)
     They heated the fire till it was very big:
     The King wanted to have a wall of iron built round those three
     who were faithful to God's Law.
(22b), 48
7)  The fire turned away from the holy men against the heathens.
6b) The three youths were blithe;
(22b). 48
7)  The fire turned against the evil men.
8)  Nebuchadnezzar saw what was happening.
6b) The three youths were blithe,
6c) they praised the Lord zealously,
6)  their lives were saved.
6c) They praised God,
     who had driven the heat of the savage fire away.*
6)  The three youths were left unharmed by the onrush of the fire,
6d) the sound of the flames worried them no more than the light of the sun.
6)  The fire did the youths no harm,
7)  but it turned against those who had done that crime, it destroyed their good looks.**
8)  Nebuchadnezzar saw what had happened.
6)  The faithful youths were walking unharmed in the oven.
8)  Nebuchadnezzar also saw one of God's angels in their company.
6d) It was inside the furnace very much as when the sun shines in summer,
      and dewfall comes in the daytime, spread about by the wind.
5) It was God who had saved the three men from the King's hatred.

*    In the Vulgate the reference is to the angel, not to God himself!
**  An effective understatement!



Word frequency and the choice of synonyms was used by Balg1, Hofer2, and Steiner3 as a criterion to decide the question of interpolation in Daniel. When we look at their findings, it becomes clear to us that among the nouns it is only the words and kennings for God that lend themselves to a conclusive investigation. There are enough instances, and it is easy to decide whether a word is a synonym for God or not. Hofer4 and Steiner5 also look into the words used for the three youths, but as Steiner himself admits, there are so few instances in Daniel B (Steiner counts three, Hofer five) that a comparison yields no conclusive results. Steiner also lists the words for "'Mann', 'Mensch', 'Menschengeschlecht', 'Volk', 'Schar'".6 His conclusions must needs be worthless because he lumps together under one heading words whose meanings are quite different.

Hofer's and Steiner's lists of words for God are reliable enough as a basis for interpretation. The mistakes, omissions and other shortcomings are not very serious. Hofer's arrangement of the list is somewhat slipshod in that it is hard to see the principle that underlies the order of the entries. Steiner's list, on the other hand, is systematic and facilitates immediate insight for the reader. The words are listed in decreasing order of frequency in Daniel A and Daniel B: nouns are put first, adjectives follow. Steiner seems to biased when he fails to list liffruma (l. 642, Daniel A 4) and lifes leohtfruma (l. 408, Daniel B3) under one heading (fruma).7 Moreover, mihtum swið (l. 283, Daniel B1) and tirum fæst (l. 311, Daniel B) should not be listed as synonyms for God9. Hofer forgets to list herra (l. 392, Daniel B3) and three instances of the word god (ll. 197, 204 and 216, all in Daniel A).

Steiner, who follows Balg, stresses the fact that Daniel A and Daniel B shore only 7 out of 68 words and kennings that are found in the whole poem.10 Hofer mentions the huge difference in the frequencies of the word god and draws our attention to words that are used mainly or exclusively in one part of the poem:
"Dan. A erwähnt die person Gottes an 96 stellen und gebraucht an 34 davon den ausdruck god; Dan.B hat dieses wort nur an einer einzigen stelle unter 39."
weard und hyrde, welche in Dan.A mehrfach vorkommen, fehlen in Dan.B ganz, dagegen hat der letztere fæder und die nomina agentis scyppend, nergend, helpend, settend, die in Dan.A nicht zu finden sind; frêa erscheint 4mal in Dan.B, 1mal in Dan.A.11

So far, so good. But in the light of the disagreement that has arisen among scholars about the question of interpolation in Daniel (its existence as well as its extent), Balg's, Hofer's and Steiner's summary conclusions will no longer do. I shall therefore begin my own investigation with a glance at the Vulgate text, and then go on to make a more detailed and also more careful analysis of the words for God in Daniel.

The words for God used in the Vulgate text are the following ones:
Dn. 1:1-3:23: Deus nine times, Dominus twice.
Dn 3:24-90 (the apocryphal section): Deus seven times, Dominus 45 times.
Dn 3:91-5:23: Deus eleven times, Dominus once, Dominator once, and Excelsus three times.

There is little variation, and that does not come as a surprise. The great variety of synonyms and kennings in Daniel is almost wholly owing to Old English poetic tradition.
Dominus, not Deus, is the commonest word in the apocryphal section. This may have prompted by the use of herra (once) and frea (four times) in Daniel B. Daniel A has only one instance of frea (in l. 651). But Deus also occurs sundry times in the apocryphal section, and the Vulgate text gives no explanation for the almost total avoidance of the word god in Daniel B. The avoidance may be partly owing to the heightened style of Daniel B. But the style of lines 232-278 (=Daniel A2) is also heightened, and the word god is found three times.

The instances of the word god are evenly spread over Daniel A; the word is used ten times in Daniel A1, three times in Daniel A2, six times in Daniel A3, and 18 times in Daniel A4. Daniel A names God 98 times and uses the word god 37 times, Daniel B names God 41 times and has the word god only once.
This discrepancy remains a mainstay in the argumentation for an interpolation in the Old English Daniel. Daniel B eschews the most straightforward word for God, but in all the sections of Daniel A, god is far more frequent than any of its synonyms.

Other heiti of generally high frequency in Old English poetry are found in similar proportions in Daniel A and Daniel B:

Daniel A
Daniel B

* metod in A2 is only found in the phrase, metodes weard.

That the proportions are a little higher in Daniel B is quite natural: god is found only once, so that drihten, metod and waldend may be said to take its place.

There are few other correspondences between Daniel A and Daniel B even if, unlike Balg, Hofer, and Steiner, we do not insist on the full likeness of these heiti and kennings:

Daniel A
Daniel B
heofona heahcyning (625, A4)
heahcyning heofones (407, B3)
witig wuldorcyning worlde and heofona (426, A3)
wereda wuldorcyning (308, B1)
rices þeoden (33, A1)
þeoden (357, B2)
liffruma (642, A4)
lifes leohtfruma (408, B3)
frea (650, A4)
frea (350, B2; 377, 395, 400, B3)
ælmihtig (272, A2; 484, A3)
ælmihtig (367, B3)

It is advisable to go beyond listing the heiti and kennings that are found only in either Daniel A or Daniel B. The question of interpolation is complicated nowadays by the question of its extent. If we look at our list of the less common correspondences between Daniel A or Daniel B, we find that the instances of likeness are spread evenly over all the sections; all of them, A1, A2, A3, A4; B1, B2, B3 have their share. There is no hint that the whole or any section of Daniel A has a particular affinity with the whole or any section of Daniel B. Thus a redefinition of Daniel A and Daniel B is not called for on this evidence.

The argument for an interpolation would become more convincing if it were possible to show that quite a few words and kennings that are peculiar to Daniel A or Daniel B are found in more than one section. If in Daniel B the correspondences were to concentrate on sections B1 and B2, the interpolation might stretch only as far as line 361, and B3 could then belong to the bulk of the poem (Daniel A). But if the correspondences are evenly spread over all the sections, then Daniel B must be regarded as a unity.

What are the correspondences within Daniel A and Daniel B?

Daniel A:

is not found in Daniel B, but in all the sections of Daniel A; this is the most striking instance of a word that is peculiar to Daniel A: this is the most striking instance of a word that is peculiar to Daniel A.

heofonrices weard (12, 26 A1)
mihtig metodes weard (235 A2)
gumena weard (237 A2)
halig heofonrices weard (458 A3)
wuldres weard (760 A4)

metodes mihte
This compound is found in line 169 (A1) and line 647 (A4).

This word is used twice in Daniel A, but exclusively in section A1 (lines 11 and 199). It is not found in Daniel B.

Daniel B:

Particularly striking is the great number of present participles (nonima agentis) other than waldend. They are found in two sections:

hæleða helpend
lean sellend
sigora settend

None of those present participles is found in Daniel A, but reccend is used in line 579 (A4).

fæder (lines 362 and 400) and herra (line 392)
are found exclusively in Daniel B, but only in section B3. This peculiarity is not enough, however, to claim that B3 is not the work of the same poet as the rest of Daniel B.

There are some words and word combinations that are found only or mainly in Daniel B, although Daniel A is over four times as long as Daniel B:

soðfæst metod
This is found in two different sections of Daniel B, lines 332 (B1) and 383 (B3), but not in Daniel A.

ece drihten
The word drihten is found in both Daniel A and Daniel B. But whereas in Daniel A the simplex word predominates (it is used in eleven out of seventeen cases), drihten stands on its own only once in Daniel B (l. 281) and is used in combination with other words eight times.
There are two instances of halig drihten (lines 292, B1, and 404, B3) against an only one in Daniel A in lines 12, one instance of witig drihten in line 403 (not found elsewhere in Daniel), and no fewer than five instances of ece drihten: two in B1, one in B2, and two again in B3. This frequency is particularly striking when we consider that ece drihten is found only twice in Daniel A (in lines 477, A3, and 716, A4). A similar, but longer phrase is in line 194-195 (A1): drihten / ece uppe ælmiht(ig)ne .

This word is found four times in Daniel B: once in B2 (l. 350), and three times in B3 (lines 377, 395, and 400), but only once in Daniel A. The frequent use of this word in Daniel B may, as has been suggested above, have been prompted by the high frequency of Dominus in the apocryphal section of the Vulgate text.


Daniel A and Daniel B share quite a few words, kennings and other compounds and word combinations for God. That is not surprising, since many of those synonyms are already found in Cædmon's Hymn, and must be looked upon as a store of poetic words and formulae that any Old English poet could draw upon. But the differences between Daniel A and Daniel B are quite great. It is true that the selection of words in Daniel A is not very profiled. Of the heiti and kennings that Daniel A does not share with Daniel B, only weard is found in all the sections.  But we must not forget to hold the very high frequency of god in Daniel A (37% of all the words for God) against the one instance of the word in Daniel B. A great contrast to Daniel B may be seen in the fact that a handful of words make up most of the instances in Daniel A. In Daniel B there is a more even distribution:

The five most frequent words for God in Daniel A (god, drihten, metod, waldend, and weard) account for 86% of all the instances, whereas the six most frequent words for God in Daniel B (drihten, metod, frea, waldend, and nergend or scyppend) make up only 59% of all the 41 instances.

Short as it is, Daniel B shows a great deal of coherence between the sections: many present participles and the combination soðfæst metod are found in more than one section, but not in Daniel A. ece dryhten is found in all the sections of Daniel B, frea in two. ece drihten and frea are frequent in Daniel B, but rare in Daniel A.

Daniel A has only twelve different words for God, Daniel B seventeen. Daniel A seeks variation in the frequent use of compounds and genitive constructions, but it uses the simplex words relatively more often than Daniel B.



Daniel A1 (lines 1-231)

god ten times (lines 21, 24, 86, 154, 156, 197, 204, 216, 219, 229)

drihten seven times (12, 32, 37, 87, 150, 194, 220)

  • halig drihten 12
  • drihtnes domas 32
  • drihten / ecne uppe, ælmihtigne 194-5
  • wereda drihtne 220

metod eight times (4, 14, 20, 36, 56, 92, 169, 174)

  • metodes mægen 4
  • metodes mægenscipe 20
  • metodes mihte 169
  • metod alwihta 14
  • metodes est 174

waldend once (wuldres waldend, line 13)

weard twice (heofonrices weard, 12 and 26)

cyning once (heah cyning, 198)

þeoden once (rices þeoden, 33)

hyrde twice (lines 11 and 199)

  • hyrde god 11
  • gasta hyrde 199

Daniel A2 (lines 232-278)

god three times (lines 236, 259, 277)

  • wuldres god 277

drihten once (257)

metod once (only in the combination mihtig metodes weard 234)

waldend once (line 240)

weard twice (234, 236)

  • mihtig metodes weard 234
  • gumena weard 236

ælmihtig once (198)

Daniel A3 (lines 409-494)

god six times (lines 421, 425, 464, 470, 473, 488)

drihten five times (438, 444, 455, 465, 476)

  • ece drihten 476

metod twice (442, 493)

  • halig metod 442
  • metod ælmihtig 493

waldend three times (lines 447, 451, 456)

  • mihta waldend 447, 451
  • rodera waldend 456

weard once (heofonrices weard, 457)

cyning once (witig wuldorcyning, 426)

dema once (dema ælmihtig, 477)

Daniel A4 (lines 495-764)

god eighteen times (lines 517, 525, 532, 548, 591, 606, 616, 618, 629, 643, 650, 669, 694, 713, 737, 742, 751, 754)

drihten five times (716, 720, 735, 744, 761)

  • ece drihten 716
  • drihtnes domas 744
  • ealra gesceafta / drihten 760-1

metod eight times (566, 578, 589, 624, 630, 647, 658, 680)

  • metodes mihte 647
  • metodes mihtum 658

waldend once (ealra gesceafta / ... waldend, 760-1)

weard once (wuldres weard, 759)

cyning once (heofona heahcyning, 625)

þeoden once (587)

liffruma once (642)

frea once (650)

ana ece gast once (626)

reccend and rice once (579)

Daniel B1 (lines 279-332)

drihten four times (lines 281, 292, 309, 330)

  • halig drihten 292
  • ece drihten 309. 330

metod twice (283, 332)

  • metod alwihta 283
  • soðfæst metod 332

waldend twice (290, 331)

  • rodora waldend 290
  • weroda waldend 331

cyning once (wereda wuldorcyning, 308)

nergend once (niða nergend 312)

scyppend twice (gasta scyppend, 291, 314)

sigora settend once (332)

Daniel B2 (lines 333-361)

drihten once (ece drihten 359)

  • halig drihten 292
  • ece drihten 309. 330

metod once (334)

waldend once (þeoda waldend, 360)

frea once (350)

þeoden once (357)

Daniel B3 (lines 362-408)

drihten four times (lines 381, 396, 398, 400)

  • ece drihten 381, 396
  • halig drihten 404
  • witig drihten 403

metod three times (383, 398, 401)

  • soðfæst metod 383
  • (soðsunu) metod(es) 401

cyning once (heahcyning heofones, 407)

ælmihtig once (heahcyning heofones, 407)

frea three times (377, 395, 400)

  • frea mihtig 377
  • liffrean 395
  • frea folca gehwæs 400

lifes leohtfruma once (line 408)

fæder twice (362, 400)

  • bylywit fæder 362
  • fæder ælmihtig 400

herra once (line 392)

halig gast once (line 402, = the Holy Ghost)

hæleða helpend once (line 402)

nergend twice (374 and 401)

  • sawla nergend (401)

scyppend twice (æhta scyppend, 291, 314)

lean sellende once (332)




The proportion of on-verses with two staves in Daniel A and Daniel B was used by Hofer1 and Steiner2 as one of their criteria in their attempt to answer the question of interpolation.
Hofer excludes from his calculation seventeen doubtful verses and thirty-eight hypermetric lines (Schwellverse) and gets the following result:

On-verses with two staves
Daniel A
Daniel B

Steiner excludes doubtful and incomplete verses, but includes hypermetric verses in his calculation. Unfortunately he does not say to which verses he takes exception. His figures are:

On-verses with two staves
Daniel A (626 verses)
44.7% (281 verses)
Daniel B (126 verses)
56.3% ( 71 verses)

Both Hofer and Steiner find that Daniel B has a higher proportion of verses with two staves than Daniel A. But their figures, especially those for Daniel B, where there are no schwellverse, are so markedly different that we must conclude that Hofer's and Steiner's definitions of double alliteration are not quite the same. It is likely that Hofer and Steiner fail to define what they mean by double alliteration because they take the definition for granted. Hofer does not even tell us whom he considers as an authority on English metre, but Steiner refers at least to the works of Schipper and Rieger3.

Hofer thinks that the difference between Daniel A and Daniel B in the proportion of verses with double alliteration is significant, but gives no further commentary. Steiner takes the same view as Hofer but he also thinks that Daniel A, with a smaller proportion of on-verses with double alliteration, must be younger than Daniel B.4 Farrell, who argues for the unity of the Old English Daniel, fails to explain why Daniel A and Daniel B differ in their handling of the metre. He does not even mention the proportions of on-verses with double alliteration.

Hofer's and Steiner's investigations are bases on obsolete scholarship: The books on Old English metre by Sievers, Heusler, Pope, Bliss and others have come out since then.5 Moreover, Hofer and Steiner do not take into account that there may be considerable differences within Daniel A and Daniel B.

I have based my own calculation of the proportion of on-verses with two staves in Daniel on the definition given by Bliss.6 I have only counted functional alliteration and disregarded ornamental and accidental alliteration.7

When we look at my results we find that the overall figures for Daniel A and Daniel B differ widely: the proportion of on-verses with two staves is 42.3% in Daniel A and 52.7% in Daniel B. But theses figures may nevertheless be inconclusive for the question of interpolation. The proportion of on-verses with double alliteration differs even more widely with Daniel A:

A1: 33.6%
A2: 76.1%
A3: 52.3%
A4: 41.1%

It may be objected that A2 and A3 contain most hypermetric verses of the poem, so that the proportion of on-verses with double alliteration must needs be higher than in the other two sections, since double alliteration is almost compulsory in hypermetric verses. But to that may be said that the high percentage of hypermetric verses itself marks off sections A2 and A3 against the other two sections. But even if we do not count the hypermetric verses in A2 and A3, we get higher proportions of on-verses with two staves than in A1 and A4. In A2 13 out of 24 non-hypermetric verses have double alliteration (54.2%), in A3 32 out of 73 (43.9%). IN A2 the proportion is thus about the same as in Daniel B (and higher than in B1).

Daniel B, on the other hand, shows greater inner consistency. The proportions are as follows:

B1: 50%
B2: 56.7%
B3: 52.7%

In the face of the divergent proportions within Daniel A it is impossible to argue for the presence of an interpolation in Daniel B on the grounds of different overall figures for Daniel A and Daniel B.

Double Alliteration in Passages of Narration, Direct Speech and Description

A comparison was made some years ago between passages of "narration", "description" and "direct speech" in the Old English Exodus.8 The percentage of on-verses with two staves is below the overall figure for the poem in passages of direct speech, but above the overall figure in passages of description.Lucas gives the following figures for Exodus:9

the whole poem (590 lines)
direct speech (77 lines)
description (41 lines)

The figures for Daniel are as follows:

the whole poem (635 lines)
direct speech (111 lines)
description (A2, 46 lines)

The figures are similar to those for Exodus, though a little lower. Daniel A2 is perhaps the nearest parallel in Daniel to the passages of description in Exodus for which the figures have been given. It differs not only in that it is a mixture of description and commentary, but also in the overwhelming share of hypermetric verses, whereas there are only a handful of hypermetric lines in Exodus, and none in the passages of description.

The direct speech in Daniel B consists of the prayers of Azarias and the Three Children. It can hardly be likened to ordinary direct speech. The link passage between the Song of Azarias and the Song of the Three Children is mainly descriptive. Its proportion of on-verses with two staves is the highest in Daniel B, though much lower than that of A2.

Appendix: The Proportion of On-Verses with Double Alliteration in Daniel

269 out of 635 verses


  78 out of 232 verses


  35 out of   46 verses


  45 out of   86 verses


111 out of 270 verses


  69 out of 131 verses


  27 out of   54 verses


  17 out of   30 verses


  25 out of   47 verses


Double alliteration in direct speech:

A1 (130-3; 136-44 = 13 lines)
  78 out of 232 verses
A2 (no direct speech)
  35 out of   46 verses
A3 (411b-5, 417b-29; 472-85 = 30 lines)
  45 out of   86 verses
A4 (551-92; 608-11; 743-64 = 68 lines)
111 out of 270 verses


Preliminary remark: Whenever figures are given in three or four columns in this chapter, the first figure gives the number of on-verses with two staves, the second the number of on-verses with one stave, the third the number of off-verses, and the fourth the percentile proportion of verses of the given type to the whole.

I shall now try to evaluate whether the statistical tables on the metre of Daniel yield any information that may help us to solve the question of interpolation.

There are marked differences in the proportions of the major verse types in Daniel A and Daniel B.

Type a

The proportion of this type is much higher in Daniel A than in Daniel B and Beowulf:

Daniel A: 9.6%
Daniel B: 6.6%
Beowulf : 5.6%

Of great interest is the fact that this type is found four times in the off-verse in Daniel A, but never in Daniel B. There are no instances in Beowulf; therefore the four off-verses of type a must be looked upon as an anomaly peculiar to Daniel A.

Type d

The proportion of this type is higher in Daniel B than in Daniel A. 77 out of 124 instances of this type are found in the on-verse in Daniel A, but only 14 out of 46 in Daniel B. The distribution in Daniel A is about the same as in Beowulf, where there are 507 on-verses out of a total of 691 verses of this type. The proportion of on-verses to the total number of verses of this type, and the proportion of d-verses to the whole are given below. The differences between Daniel A and Daniel B are remarkable.

Proportion of on-verses
Proportion of d-verses to the whole

Daniel A


Daniel B





The share of 1A-verses is higher in Daniel A than in Daniel B, but similar to that in Beowulf. The figures below give us the proportion of 1A-verses to the whole:

Daniel A: 6.1%
Daniel B: 4.3%
Beowulf : 6.0%

(Type 1A2 occurs only twice in Daniel, once in Daniel A, and once in Daniel B. In Beowulf, 9% of all the 1A-verses belong to this type.)


The proportion of 1A*-verses is higher in Daniel B than in Daniel A and Beowulf. In Daniel A, as in Beowulf, about half t verses of this type are in the on-verse, and most verses of this type in the on-verse have double alliteration. In Daniel B the proportion of on-verses with double alliteration is about the same as in Daniel A and Beowulf, but the proportion of on-verses is much higher than in Daniel A and Beowulf.


2 staves in on-verse
1 stave in on-verse
Proportion of this type to the whole
Daniel A
Daniel B


The proportion of verses of this type is higher in Beowulf than in Daniel B. It is lowest in Daniel A.


2 staves in on-verse
1 stave in on-verse
Proportion of this type to the whole
Daniel A


Daniel B




We can see from the table above that there is a difference in the distribution of 1D-verses between Beowulf and Daniel A on the one hand, and Daniel B on the other hand. But it must not be overlooked that the figures for Daniel B are rather small and should be compared with caution.


The distribution is as follows:


2 staves in on-verse
1 stave in on-verse
Proportion of this type to the whole
Daniel A


Daniel B




The proportion of 1D*-verses is about the same in Beowulf and Daniel B, but lower in Daniel A. The two instances of type 1D* in the on-verse with single alliteration are highly irregular. No parallel can be found in Beowulf and Daniel B, and we may say that Daniel A, as in the case of type a, breaks a rule that can be observed in Beowulf, whereas Daniel B conforms.


The proportions are similar:

Daniel A: 17.6%
Daniel B: 20.2%
Beowulf : 19.9%

In Beowulf, slightly more than half the verses of type 2A are found in the on-verse, and of these again slightly more than one half have double alliteration. In Daniel A and Daniel B only about two fifths of the verses of this type are in the first half-line. But whereas the proportion of on-verses with two staves is not much higher in Daniel than in Beowulf, three out of four on-verses in Daniel B have double alliteration!


2 staves in on-verse
1 stave in on-verse
Proportion of this type to the whole
Daniel A


Daniel B





The proportion of verses of this type is about the same in Daniel A as in Beowulf, and not much lower in Daniel B. The number of on-verses of this type is always small, and verses with two staves are rare:

Daniel A: 3.2% of all the staves
Daniel B: 2.3%
Beowulf : 2.9%


The share of verses of this type is biggest in Daniel A and smallest in Daniel B. The differences are great:

Daniel A: 11.3%
Daniel B:  6.2%
Beowulf :  8.5%

Most instances of this type are found in the off-verse, and in the case of Daniel B the number of on-verses of this type is so small that a comparison of the proportions of on-verses with double alliteration to the total number of on-verses would be inconclusive.


The proportion is about the same in Beowulf and Daniel A, but higher in Daniel B:


2 staves in on-verse
1 stave in on-verse
Proportion of this type to the whole
Daniel A


Daniel B




The proportion of 1D*-verses is about the same in Beowulf and Daniel B, but lower in Daniel A. The two instances of type 1D* in the on-verse with single alliteration are highly irregular. No parallel can be found in Beowulf and Daniel B, and we may say that Daniel A, as in the case of type a, breaks a rule that can be observed in Beowulf, whereas Daniel B conforms.


The distribution in Beowulf and Daniel B is about the same, but the comparison is inconclusive because of the small number of such verses in Daniel B. In Daniel A half the verses of this type are in the first half-line, but only two out twenty have two staves. However, this disproportionate distribution is largely due to the nine instances of on-verses of type 3B* with single alliteration in section A1!


2 staves in on-verse
1 stave in on-verse
Proportion of this type to the whole
Daniel A


Daniel B





Verses of this type are almost twice as frequent in Beowulf as in Daniel A and Daniel B. The figures are:

Daniel A: 3.8%
Daniel B: 3.1%
Beowulf : 5.9%

Seven out of eight verses in Daniel B are in the first line, whereas in Daniel A 29 out of 48 instances are in the on-verse. In Beowulf more than two thirds of the verses are in the off-verse. So in this case Daniel A and Daniel B seem to agree against Beowulf, but it must be borne in mind that the figures for Daniel B are rather small.

Hypermetric verses

Hypermetric verses are scattered over all the sections of Daniel A, but none are found in Daniel B. There are clusters of hypermetric lines in the furnace episode (lines 224-278) and in lines 430-475, which sum up the outcome of the furnace episode and mark the conversion of Nebuchadnezzar. It seems that the function of hypermetric verses is to heighten the style of the poem.2 There are no hypermetric verses in Daniel B albeit the style is heightened throughout, but the proportion of normal on-verses with double alliteration is higher than in Daniel A. It would be well to bear in mind that A2 (lines 232-278) and B2 (lines 333-361) tell the same episode, but are metrically very different.


The difference in the proportions of the various verse types between Daniel A and Daniel B is remarkable even though the statistical divergences are in part interdependent. Especially striking is the absence of any hypermetric lines from Daniel B.

The Relevance of the Differences in the Distribution of the Verse Types

If these differences betoken that Daniel A and Daniel B are not the work of the same poet, we may expect that the metrical differences between the four sections of Daniel A are smaller than between Daniel A and Daniel B. This is indeed borne out by a detailed investigation! The table below shows the proportions of the major verse-types for A1, A2, A3, A4, and B. So we get five figures for each type, and if we number these figures in the order of the proportions, we find that the figure for Daniel B is often an extreme one. Sometimes it is the highest but one or lowest but one, but never the one in the middle. If we do not include in the comparison percentile figures that are based on very small absolute numbers (one, two, or three), then the agreement between the sections of Daniel A against Daniel B is even more striking.

The exclusion of hypermetrical verses from the calculations underlying the statistical table is required because the figures for A2, and to a lesser degree those for A3, would be distorted, since the major clusters of hypermetric verses are found in these sections. In A2 they make up 43.3% of all the verses!



The Proportions of the major Verse Types in Daniel to the Total Number of Verses (Excluding Hypermetric Lines)

Verse Type
DANIEL A1 (435)
DANIEL A2 (51)
DANIEL A3 (140)
DANIEL A4 (534)
DANIEL B (258)
13.1 (57)
7.8 (4)
9.0 (13)
8.8 (47)
6.6 (17)
11.7 (51)
3.9 (2)
5.0 (07)
13.5 (72)
17.8 (46)
5.7 (25)
7.8 (4)
8.6 (12)
6.7 (36)

4.3 (11)

20.9 (91)
15.7 (8)
19.3 (27)
18.0 (96)
20.2 (52)
6.7 (36)
11.8 (5)
3.6 (05)
7.7 (41)
10.1 (26)
10.1 (44)
9.8 (5)
12.8 (18)
10.5 (56)
12.0 (31)
10.8 (47)
9.8 (5)
13.6 (19)
13.5 (72)
6.2 (16)
3.4 (15)
3.9 (2)
5.0 (07)
3.7 (20)
4.6 (12)
3.4 (18)
2.0 (1)
5.7 (08)
3.9 (21)
3.1 (08)
The Proportions of On-Verses to the Total Number of Verses of the Given Type (Excluding Hypermetric Verses)
62.8 (32)
(50 (1))
85.7 (6)
52.8 (38)
30.4 (14)
The Proportions of On-Verses with Double Alliteration to the Total of On-Verses (Excluding Hypermetric Verses)
50 (17)
(100 (1))
45.6 (5)
59.5 (25)
75 (15)

Daniel B has the highest or lowest percentage in each category, or the highest or lowest but one. The figure in B is in bold type and underlined if it is extreme, if it is the one nearest to the extreme it is in bold type and italics. The figure in the section of Daniel A that is nearest to the value in Daniel B is marked in the same way. Percentages based on very small absolute numbers (1 or 2) are in smaller print.



There are many verses in Daniel that do not conform to the standard of Beowulf. These verses in Daniel fall into the following categories:

01) Verses that are metrically trisyllabic
02) Verses with postponed alliteration
03) Verses of type a in the second half-line
04) Lines without alliteration
05) On-verses that have only one stave although they ought to have double alliteration
06) Off-verses of types where double alliteration is compulsory or almost compulsory, and which therefore ought to be found only in the first half-line
07) Double alliteration in the off-verse
08) Stress and stave on words that are neither stresswords nor displaced from their normal positions in the clause
09) Lack of stress on words that ought to take stress
10) Anomalous word order
11) Verses that do not scan at all
12) Verses of types that are extremely rare in Beowulf
13) Verses that are almost but not exactly paralleled in Beowulf

It is hard in some cases to draw the line between anomalous and rare or doubtful verses. Metrically trisyllabic verses are so rare in Beowulf that they must be looked upon as anomalous. Other types in Daniel that are extremely rare in Beowulf or that are not even exactly paralleled in Beowulf, although they may be seen as variants of existing types, have been reckoned among the rare and doubtful verses. Some of them may in fact be anomalous.

The many instances of verses in which the possessive adjective bears stress and stave are a special case. This phenomenon is rife in Daniel, but not elsewhere in Old English poetry. Some of the instances in Daniel have exact parallels in other poems, but the possessive adjective sin is regularly stressed in Daniel when it stands before the noun, and this usage is peculiar to this poem. The normal pattern is found in lines 126 (swefnes sines) and 135 (witgum sinum), where sin takes stress because it is displaced from its normal position before the noun it qualifies. But much more frequent in Daniel is sin bearing stress and stave in its normal position. We can speak of a systematic metrical idiosyncrasy in Daniel. The striking thing about this peculiar metrical treatment of sin in Daniel is that it is not paralleled in the handling of min and þin. The metrical handling of possessive adjectives in Daniel is systematic but capricious.

We can only guess and speculate about how sin came to be treated in Daniel like a stressword. The normal position of min in Beowulf is before the noun that it qualifies, and min is then seldom stressed. Occasionally min follows the noun it qualifies. The following examples are particularly interesting:

wine min Beowulf 457b, 1704b
wine min Unferð 530b

Bliss lists these verses as heavy verses in which one of three stressed elements must be subordinated.1That is consistent with the rule that articles and proclitics are stressed when they are displaced from their normal position in the sentence.2 But a different explanation may be possible which, if it should be borne out by the facts, would complicate things a little.

In Old Norse (as in modern Icelandic and Swedish dialects) the possessive adjective can forego or follow the noun it qualifies, and both positions are normal with only exceptional stress on the possessive. It seems to me not impossible that traces of this free treatment of the possessive are still found in Old English poetry. Verses 457b, 1704b, and 530b would then be normal verses with two stresses, not heavy verses. If the possessive can be unstressed in either position, before or after the noun, there is no clear guidance as to when the possessive should take positional stress. A poet would thus be free to stress on of the possessives systematically in both positions, and that may have happened to sin in Daniel. The verses where sin bears stress and stave although it stands before the noun are then not anomalous or very rare, but they are peculiar to Daniel as a group because of the regularity of the phenomenon.

Only one verse in Daniel B is anomalous, and another doubtful. 4.3% of all the verses in Daniel A fail to conform to the standard of Beowulf, but only 0.8% in Daniel B. The difference is very marked and makes it likely that Daniel B is an interpolation.3


(The numbering refers to the categories given at the beginning of this section)

Anomalous verses and lines in Daniel A:
01) 159 a
02) 92a, 202b, 460b
03)122b, 138b, 410b, 529b
04) 172, 192, 266, 449, 710
05) 174a, 224c, 532a, 640a, 742a
06) 58b, 229b
07) 91b, 608b
08) 137b, 473b
09) 440b
10) 590b, 644b
11) 164a, 180b, 596a, 707a, 745a

Verses in Daniel A that are doubtful or belong to types that are rare:
01) 196b
06) 46b
08) 416a
11) 621a
12) 539a, 549a
13) 554a, 576a, 646a

Verses in Daniel A where the possessive sin bears stress and stave when it stands before the noun:
08) 75b, 79b, 108b, 210b, 159b, (449a), 468b, 526b, 648b, 758b
Verse 449a is also irregular in that it shares no stave with the off-verse. The line is therefore reckoned among those lines in Daniel A that lack alliteration (category 4).



There are some instances of ornamental end-rime in Daniel. End-rime is more frequent than in Exodus,1 but still by no means accidental. A list of end-rimes is given by Steiner:2

101b:103a    dæde: wæde
104b:105a    weard: middangeard
214b:216a    wolde: golde
523b:525b    ende: sende
567a:569a    aceorfeð: onhweorfeð
606a:607a    sealde: gewalde
635a:636a    weard: middangeard
649b:651b    ateah: beseah

The last example is not a pure rime. the diphthong in ateah is long, that in beseah is short. Spot-checks reveal that Steiner does not give a full list. His second example is in fact an instance of triple rime:

212b:214b:216a    scolde: wolde: golde

Further examples are the following ones:

200b:202b    cwædon: gebædon
576a:576b    weceð and wreceð

Steiner also gives a list of impure rimes ("ungenaue Reime"):

   47b:48b    burhstede: niðhete
546a:547a    witgode: swigode
640b:641b    com: aldordom
695a:698b    grome: cwome

To these impure rimes might be added those instances where the stressed syllables rime, but the endings are different. I have found the following examples:

   14a:14b    wolde: golde
256b:258b    heredon: generede
476b:477a    drihten: ælmihtig
513a:513b    besnædan: blædum
753a:753b    aldor: wolde

There may originally have been identical rime in the following case:

166b:168b    onfon ne meahte: gedon mihte

These examples of pure and impure rime are scattered all over Daniel A, but none are found in Daniel B. The absence of end-rime in Daniel B is not in itself a clear token of interpolation, but it confirms earlier findings and gives thus more weight to the argumentation for an interpolation.

End-rime becomes on the whole more frequent towards the end of the Old English period than in the early poetry.3 We must reckon with individual choice or avoidance of endocrine, of course, but if we think of other features that point to a relatively late date of composition, e.g. the number of lines without alliteration and generally the lesser degree of conformity with the metre of Beowulf, we may deem it likely that Daniel A was written later than Daniel B.

Appendices to Chapter III:

1) Index to the Scansion of Daniel

1) Notes on the Scansion of Daniel

3) The Scansion of Biblical Names in Daniel

4) Detailed Tables of the Distribution of the Verse Types in Daniel


The choice of criteria to decide questions of interpolation is always a tricky business. We must reckon with the efforts of an interpolator to harmonize the style and contents of an interpolation with the rest of the text.* We must also take into consideration that the nature and contents of a supposed interpolation can be so different from the rest of the text that many criteria simply cannot be applied. Daniel B is a poetic paraphrase of apocryphal matter that is of a predominantly lyrical nature, whereas Daniel A is an adaptation of those chapters of the Book of Daniel that have a primarily narrative function. A further difficulty is the shortness of Daniel B. Many criteria that would be helpful if the supposed interpolation were longer, cannot be applied, because there is so little statistical material that the results of an investigation would not be significant.

I have tried to shun all the pitfalls by choosing what I look upon as the best criteria. I think, for instance, that the proportions of end-stopped and run-on lines yield no conclusive evidence because of the different functions of Daniel A and Daniel B. My study is by no means exhaustive, of course, as can easily be seen from a comparison of my thesis with the works of Hofer and Steiner. But I have tried to apply my criteria with great care, and I hope that I have given enough evidence to prove that there is indeed an interpolation in Daniel. I also hold that my findings make it very likely that the interpolation stretches as far as line 408, although the evidence for this is not equally conclusive.

What has remained outside this investigation is the question of the relation between Daniel B and Azarias. Such an investigation would call for a rather different methodical approach, although a consideration of the metre might again be helpful. I do not think that a study of the relation between Daniel B and Azarias would be very promising without a firm stand on the question of interpolation in Daniel.

* On this problem see for instance McTurk, p. 110, on the question of interpolation in Fóstbræðra saga: Jónas Kristjánsson's examples of the learned style in the saga are "in some cases located so close to the passages traditionally regarded as digressions, that it is difficult to exclude the posssibility of their inclusion in the text having originated in the efforts of an interpolator to harmoniye the style of the digressions with that of the remainder of the saga".




1) A similar device is found in Hallberg's study. he divides Heimskringla into two sections so as to be in a position to check his results.

Chapter 1
  1) Steiner, p.5-42.
  2) Steiner, p.21-5.
  3) Hofer, p.169-175.
  4) Farrell, p.22-9, especially p.26 and 28f.
  5) See Farrell , p.24, and Lucas, p.52!
  6) Farrell, p.29.
  7) Pope, p.41 (footnote 51).
  8) This survey is largely based on the line-for-line comparison by Steiner (Steiner, p. 5-42), who is in turn indebted to A. Ebert, Allgemeine Geschichte der Literatur des Mittelalters im Abendland, Leipzig 1875-87 (3 vls.), especially Vol.3, p.21ff.
  9) Steiner, p.12-13.
10) Steiner, p.13-14.
11) Steiner remarks: "Es kann hier auch gleich bemerkt werden, dass Zusätze psychologischer Natur, zu denen zum Teil die Quelle den Anlass gegeben hat, auch sonst sich mehrfach in unserer Dichtung finden." (p.16)
12) Steiner says: "Alles Nebensächliche, was nicht zur weiteren Entwicklung der Handlung beiträgt, Namen untergeordneter Personen und Zahlenangaben werden weggelassen." (p.10)
13) Farrell, p.26.
14) Farrell, p.25-26, p.29.
15) Pope, p. 42 (footnote 51).
16) Verses in the Vulgate text sometimes contain two sentences. In those cases I call the first part of the Bible verse a, the second b, e.g. 19a and 19b.

Chapter 2
  1) Balg, p.29-37.
  2) Hofer, p.164-8.
  3) Steiner, p.74-83.
  4) Hofer, p.166-7.
  5) Steiner, p.78-9.
  6) Steiner, p.79-82.
  7) Steiner, p.76-7.
  8) Steiner, p.77.
  9) Hofer, p.164-6.
10) Balg, p.30-1, and Steiner, p.77.
11) Hofer, p.166.

Chapter 3, Section 1
  1) Hofer, p.164.
  2) Steiner, p.42.
  3) Steiner, p.42-3 (including footnotes p.43).
  4) Steiner, p.43.
  5) The works to which I have referred are listed in the "Bibliography and List of Abbreviations".
  6) Bliss, §6 (p.4), fn.1.
  7) See Bliss, §15 (p.12) and §21 (p.17).
  8) C. Wall, 'Stylistic Variation in the OE Exodus', ELN VI (1968), p.79-84, esp. p.83. The essay is briefly discussed by Lucas (Lucas, p.42).
  9) Lucas, p.42.

Chapter 3, Section 2
  1) The figures for Beowulf are all taken from Bliss, p.122-3 and p.128.
  2) For general discussions of the functions of hypermetric verses see A.C. Bartlett, The Larger Rhetorical Patterns of Anglo-Saxon Poetry, New York 1935, and B.J. Timmer, 'Expanded lines in Old English poetry' in Neophilologus XXXV (1952), p.226-230.

Chapter 3, Section 3
  1) Bliss, §79 (p.72).
  2) The rules governing word stress in Old English were formulated by Hans Kuhn in 'Wortstellung und -betonung im Altgermanischen' in PBB LVII (1933), p.1-109. A wording of the rules in English is found in Bliss, §9 (p.6-7).
  3) Farrell also mentions irregular verses and verses of types that are rare (Farrell, p.21-22). His remarks are rather sketchy, his lists incomplete, his classification of these verses not made carefully enough, and there are obvious mistakes. It is easy to trace these shortcomings when my own list is consulted. But even Farrell ought to have noticed that nearly all the anomalies he mentions are found in Daniel A, and very few in Daniel B. Some of his statements (under the heading "Comparison of metrical uses in 'Daniel' and 'Azarias', p.22) show Farrell's awareness of the fact that hypermetric lines and some metrical irregularities are found only in Daniel A. But he fails ato say how these differences in the metre are to be accounted for if the unity of Daniel is assumed.

Chapter 3, Section 4
  1) On end-rimes in Exodus see Lucas, p.43!
  2) Steiner, p.53.
  3) Kluge, p.449, quoted by Steiner, p.54. See also Eduard Sievers, 'Old Germanic Metrics and Old English Metrics' in Essential Articles for the Study of Old English Poetry, ed. J.B. Bessinger & S.J. Kahrl, Hamden, Conn. 1968, p.267-88.

Appendix 1 to Chapter 3
  1) A detailed analysis, e.g. 2A1b instead of 2A1, is given only occasionally.

Appendix 2 to chapter 3
  1) References to the occurrence, frequency and distribution of verse types in Beowulf are always based on Bliss, esp. p.122-8.
  2) Bliss, §60 (p.53-4).
  3) Farrell, p.22.
  4) See Bliss, §44 (p.38).
  5) See Bliss, §§46-7 (p.40-2), and Lucas, p.40!
  6) See Bliss, Half-lines, p.446!
  7) Bliss, Half-lines, p.446.
  8) See Bliss, p.162-8 ("Index to the Scansion of the Hypermetric Verses in Old English")!
  9) Bliss, p.167.
10) Bliss, Half-lines, p.446.
11) Bliss, §29 (p.23).
12) Bliss, §13 (p.10).
13) Bliss, §§46-7 (p.40-2). Stanley seems to have overlooked this instance of an unstressed infinitive, for it is missing on his list (Stanley, p.324). It is interesting to see that secan in Daniel 440b behaves in a different way from secan in Beowulf 756a and 821a. In the two instances in Beowulf unstressed secan follows another infinitive, fleon, in apposition, and fleon is in each case directly dependent on a finite verb: wolde in the first instance, and scolde in the second instance. In Daniel 440b, secan is directly dependent on a finite verb, gewat. It is not followed by another infinitive in apposition or coordinated to it by a copulative. So secan in Beowulf follows the second rule, secan in Daniel the first rule worded by Stanley in his article (Stanley, p.322-3).
14) On elision in Old English see Sievers, p.890, and Bliss, e.g. §§81-2 (p.73-5).
15) Sievers-Brunner, §149 (p.126-7).
16) Professor Bliss drew my attention to this recognized weakening of wilddeor.

17) See Appendix 3 to Chapter 3!
18) This addendum goes back to a hint given to me by Professor Bliss.
19) Here I am likewise indebted to Professor Bliss.

Appendix 3 to Chapter 3
  All the references are in the text.

Appendix 4 to Chapter 3
  The order and the numbering of the verse types are the same as in Bliss' Table 1 in Appendix C (Bliss, p.122-3).