Manuscripts containing Krákumál
Krákumál ‘Speeches of the Crow’ (Anon Krm) is a monologue placed in the mouth of the legendary viking hero, Ragnarr loðbrók, as he dies in a snake-pit, the victim of the Anglo-Saxon King Ælle (ON Ella, see sts 24, 26-7 below and cf. Ragn 26-7). It is preserved in two main mss: NKS 1824 b 4° (1824b) of c. 1400 and UppsUB R 702ˣ (R702ˣ) of c. 1600-1650. In 1824b it is preserved incomplete, lacking st. 16 and breaking off halfway through l. 5 of st. 22. In R702ˣ it is preserved in twenty-nine stanzas, apparently its full complement. The text of the poem in the skaldic anthology R702ˣ is a copy made by Magnús Ólafsson of Laufás (d. 1636) from a vellum ms. now lost, and a second copy of this same lost vellum, also by Magnús Ólafsson, forms the basis of the text of the poem as edited by Ole Worm (d. 1654), printed in his Runer seu Danica Literatura Antiquissima … (1636) (LR), and consisting of twenty-nine stanzas. This second copy by Magnús is no longer extant, but a copy of it, made by Stephanus Stephanius (d. 1650) some time after 1635, survives in UppsUB R 693ˣ (R693ˣ); see Jakob Benediktsson (1948, 223-8, 343, 460-3, 515) and Faulkes (1993a, 104-7). This ms., which is apparently an accurate copy of Magnús’s work (Faulkes 1993a, 106), follows each stanza of Krm with a translation into Danish and explanatory notes in Latin. A version of these notes, edited from Magnús’s lost second copy, accompanies the text of Krm in LR, where a line-by-line Latin translation of the poem, based on Magnús’s Danish translation, is also given. There is some variation in wording among the texts of the poem as they appear in R702ˣ, LR and R693ˣ. A fourth ms., AM 6 folˣ (6ˣ) of c. 1600-1700, preserves a copy of the poem, also consisting of twenty-nine stanzas, made by Jón Erlendsson (d. 1672), apparently on the basis of 1824b and R702ˣ, and with marginal readings supplied from LR; it is doubtful how far this ms. can be said to have independent value. Finally, AM 147 4° (147), of c. 1400-1500, known as Heynesbók, preserves the poem only fragmentarily. In 1824b and 147 the text of the poem is written as continuous prose; in R702ˣ each stanza forms a separate prose paragraph; and in LR, R693ˣ and 6ˣ the poem is presented as verse, in a sequence of stanzas, as in modern editions. These five mss and the first edition (1636) of LR are used in the present edition.
In each of 1824b and 6ˣ the text of Krm immediately follows that of Ragn, forming a kind of appendix to it, while in each of R702ˣ, LR and R693ˣ, none of which contains a text of Ragn, it appears as a distinctly separate item. In 147, on the other hand, it appears within the text of Ragn, where it is presented as spoken by Ragnarr in the snake-pit shortly before he utters Ragn 26-7. The text of Ragn in 147, which thus includes that of Krm, is overwritten, and only scattered portions of it can now be read. In his edition of Vǫls and Ragn, Olsen (Ragn 1906-8, 187-9) printed all that he was able to read of Krm in 147, i.e. fragments of sts 1, 6-9, 12-15, 17-29. Soffía Guðný Guðmundsdóttir, who has recently subjected the stanzas of Ragn and Krm in 147 to careful scrutiny, has been able to read a greater number of the Krm stanzas than Olsen could (see st. 11, Note to [All]), and she has added considerably to his readings of the stanzas accessible to him. She also acknowledges the possibility that st. 16, which is lacking in 1824b, and which Olsen believed was also lacking in 147 (Ragn 1906-08, 187-8), should be counted among those stanzas of Krm in 147 that cannot now be read (see st. 16 Note to [All] below). The present editor is heavily indebted to Soffía for her readings of the text of Krm in 147, which are used in the present edition. Readings from LR are transcribed in the present edition, in lower-case roman letters, from the runic script of the younger fuþark in which Worm printed the poem, evidently believing that this was the script in which it was first committed to writing (cf. Heinrichs 1978, 294‑5).
It was mentioned above that in 147 Krm appears within the text of Ragn, placed in the mouth of Ragnarr as he dies in the snake-pit, and that in 1824b it follows the text of Ragn as a kind of appendix. Its narrative content also differs markedly from that of Ragn, as well as from those of RagnSon and Saxo. In 147, which preserves the X redaction of Ragn (see Introduction to Ragn), the inclusion of Krm as part of the action of the saga is decidedly awkward: not only does it refer retrospectively to many of Ragnarr’s exploits that are nowhere related in Ragn, but it also appears to be followed in the prose text by an announcement of his death, and shortly afterwards by the presentation of him as speaking two additional stanzas Ragn 26-7 (Ragn 1906-8, xcii, 189). The likelihood is that the compiler of the Y redaction of Ragn, preserved in 1824b, noticed these discrepancies and relegated the poem to the status of appendix that it seems to have in that ms. (see McTurk 1975, 62 n. 70).
Mss containing Krm that have not been used in the present edition are now listed. Mss of Ragn that contain Krm either complete or in part, whether preceding or following the text of Ragn or forming part of it, include the following: Borgarnes, Héraðskjalasafn Borgarfjarðar: Sagnahandrit Jóhannesar Jónssonarˣ; Cambridge, MA, Harvard Houghton Icel 32ˣ; Copenhagen AM 2 folˣ, AM 3 folˣ, AM 6 folˣ, AM 7 folˣ and GKS 1006 folˣ; Dublin TCD 993ˣ; Edinburgh Adv 21 5 2ˣ; London BLAdd 24969ˣ, BLAdd 11160ˣ; Paris, BSG: ms. 3714ˣ; Reykjavík ÍB 76 4° Iˣ, JS 8 folˣ, JS 12 folˣ, Lbs 170 folˣ, Lbs 272 folˣ, Lbs 361 4°ˣ, Lbs 1246 4°ˣ, Lbs 1487 4°ˣ, Lbs 1491 4°ˣ, Lbs 4661 4°ˣ (Lbs 824 4°ˣ and Lbs 2341 4°ˣ each refer to Krm, at the end of Ragn and within its text respectively, but neither preserves in its present state any part of the poem); Stockholm Engestr B III 1 19ˣ, Holm papp 38 4°ˣ, Holm papp 95 folˣ; Uppsala Ihre 77ˣ, UppsUB R 703ˣ, Westin 89ˣ. Of these mss those in Borgarnes, Cambridge MA, Dublin, Edinburgh, London, Reykjavík, Stockholm and Uppsala, and the Copenhagen ms. GKS 1006 folˣ, have been personally inspected by the present editor, either at first hand or on the internet, and appear to have no independent value; this assumption is made of the remaining mss listed in this and the next two paragraphs for the purposes of the present edition.
Mss of Ragn not as yet checked for text of Krm by the present editor are as follows: Berlin Germ qu. 935ˣ (copy of Björner’s edition, see below); Copenhagen AM 5 folˣ, AM 8 folˣ, AM 281 4°ˣ, AM 282 4°ˣ, AM 762 4°ˣ, AM 930 4°ˣ, Rask 32ˣ and Kall 616 4°ˣ, NKS 1144 folˣ, NKS 1157 folˣ, NKS 909 4°ˣ; London BLAdd 6121ˣ, BLAdd 11174ˣ; OsloUB 1157 8°ˣ; Rostock, Universitätsbibliothek: mss phil. 78/2ˣ; Stockholm Holm papp 21 folˣ; Trondheim DKNVSB 356 4°ˣ.
Mss in which Krm appears as a separate item, whether complete or in part, include the following: Copenhagen NKS 1705 4°ˣ, NKS 1892 4°ˣ, Thott 1497 4°ˣ, Thott 1500 4°ˣ; Dublin TCD 1016ˣ; London BLAdd 11173ˣ; Reykjavík AM 738 4°ˣ, AM 761a 4°ˣ, AM 761b 4°ˣ, Íb 300 4°ˣ, Lbs 214 4°ˣ, Lbs 636 4°ˣ, Lbs 756 4°ˣ, Lbs 1689 4°ˣ, Lbs 1849 8°ˣ; Uppsala UppsUB R 692ˣ. Of these only the Reykjavík mss listed have so far been inspected by the present editor.
The title Krákumál
Kålund (1900, 236) could read the title Krákumál (‘Krakv mal’) at the end of the line immediately preceding that on which the poem begins in 1824b (fol. 79r), but it is no longer fully legible. In 6ˣ the poem has the heading: Krákumál, er sumir kalla Loðbrókarkviðu ‘Krákumál, which some people call Loðbrókarkviða’. The state of 147 makes it impossible to say whether a title is applied to the poem in that ms. In R702ˣ and LR the poem is headed Bjarkamál ‘The Speeches of Bjarki’; in R702ˣ the heading reads: Úr Bjarkamálum er orti Ragnarr kóngr ‘From Bjarkamál, which King Ragnarr composed’; in LR its wording is Bjarkamál, sem orti Ragnarr loðbrók ‘Bjarkamál, which Ragnarr loðbrók composed’; so also R693ˣ.
The application of the title Bjarkamál to the poem is clearly an error, perhaps due to an imperfect recollection of a tradition concerning the heroic death of one Agnerus/Agnarr, a namesake of the Agnarr mentioned in Krm 17/7. This tradition is preserved in the poem in Latin hexameters by Saxo Grammaticus that is now known as Bjarkamál (Saxo 2015, I, ii. 7. 4-28, pp. 122-41), which evidently included the Old Norse Bjarkamál in fornu ‘The Old Speeches of Bjarki’ (Anon BjarkIII) among its sources (Friis-Jensen 1987, 70, 152, 178). In Saxo’s poem one of its three speakers, Biarco (= Bjarki), relates that he slew this Agnerus, son of Ingellus (= Ingjaldr), and that Agnerus died laughing, showing great courage in the face of death (Saxo 2015, I, ii. 7. 19, pp. 134-7). This recalls the final line of Krm: læjandi skal ek deyja ‘I’ll die laughing’. The tradition of this Agnerus/Agnarr dying smiling or laughing is also recorded in the prose of Saxo’s account (Saxo 2015, I, ii. 6. 10, pp. 116-19), and in the anonymous Icelandic Bjarkarímur, dating from c. 1400 (Finnur Jónsson 1904b, xxx, 161). The phrasing Úr Bjarkamálum ‘From Bjarkamál’ in R702ˣ might give the impression that Krm originally consisted of more than the twenty-nine stanzas found in this ms. as well as in LR, R693ˣ, and 6ˣ, but there is no reason to suppose this: Magnús Ólafsson is likely to have been no more accurately informed in this respect than he seems to have been about the poem’s title.
If Krákumál is indeed the poem’s earliest known title, what does it mean and how is it to be explained? In Ragn Kráka ‘Crow’ is the disrespectful nickname given to Ragnarr’s second wife Áslaug, the daughter of Sigurðr and Brynhildr, by the farming couple who bring her up as a drudge (see Introduction to Ragn). The nickname and the story attached to it seem, however, to belong to a relatively late stage in the development of traditions about Ragnarr loðbrók; they are found neither in RagnSon, behind which may lie a lost redaction of Ragn, nor in Saxo’s account of Regnerus Lothbrog in Book IX of his Gesta Danorum, composed most probably c. 1200, nor in Krm, which seems to reflect a relatively early stage in the development of traditions of Ragnarr loðbrók; it mentions Áslaug (st. 26/3) but not Kráka: Saxo mentions neither Áslaug nor Kráka.
The difficulty of explaining the title Krákumál by reference to Ragnarr’s wife Áslaug-Kráka is illustrated by an argument put forward by de Vries (1928c, 123-6), and later summarily abandoned by him (de Vries 1964-7, II, 40 n. 71). He claimed that sts 1-21 reflect the original Krákumál, or ‘Words of Kráka’, a poem placed in the mouth of Ragnarr’s wife Áslaug-Kráka, while sts 22-9 were originally a separate poem, a Ragnarsmál or ‘Words of Ragnarr’, presented as spoken by Ragnarr as he dies in the snake-pit. There is no evidence internal to the poem to support this view.
A more acceptable argument is that of Olsen (1935), whose discussion takes account of the fact that kráka as a common noun means ‘crow’. He takes the phrase í sjálfum Suðreyjum ‘in the Hebrides themselves’ in Krm 15/3 (see Note there), with its definitive adjective sjálfum, as a strong indication that the poet of Krm lived in the Hebrides, or had close ties with them, and that it was in the Norse-Celtic environment of the Hebrides that the poem was composed. He further maintained that the poem’s title, Krákumál, lit. ‘Crow’s words’ or ‘Things told of a crow’, is a Norse calque of Hebridean origin on OIr. badbscél ‘tale of slaughter’, the two elements in which, badb and scél, mean respectively ‘crow’ and ‘story’; badb often refers in Old Irish literature to a Celtic war-goddess who appears in battle-scenes in the form of a crow. This view makes Krákumál ‘Tale of the Crow’, understood as ‘Tale of Slaughter’, an eminently suitable title for a poem dealing with battles of one kind or another from beginning to end, and encourages the further view (eventually accepted by de Vries, 1964-7, II, 40 n. 71) that Krm is a unity and has been so from the start. The former view also implies that the title originally had nothing to do with Kráka ‘Crow’ applied as a nickname to Ragnarr’s wife Áslaug. Olsen (1935, 80) indeed hints strongly that the title Krákumál, no longer understood in its original meaning once the poem had come to be known in an exclusively Scandinavian (as opposed to Norse-Celtic) environment, may have stimulated the influence on Ragn of Saxo’s story of Craca, stepmother of Ericus Disertus ‘Eric the Eloquent’ (Saxo 2015, I, v. 2. 3, pp. 264-5); see further McTurk (1991a, 118-36; 2012b, 373‑6; 2014a, 57‑8).
Date and place of composition
The remarks above imply that Krm was composed before c. 1250, the probable date of the X redaction of Ragn in which Áslaug’s nickname Kráka first appears. Storm (1878, 196-200) argued that Krm was originally composed in Denmark in the twelfth century, and was later reworked in Iceland, where it was given its present and, as he believed, augmented form no earlier than the second half of the thirteenth century. His reasons include the closeness of its narrative content, as he saw it, to Saxo’s account of Regnerus Lothbrog; the fact that its spelling shows some instances of the lack of initial <h> before <l> and <r>, reflecting a sound change not characteristic of Old Icelandic (see however ANG §289 Anm. 2); and the lateness of its vocabulary. It is not clear whether he saw the poem’s mention of Áslaug as one of the additions made at the Icelandic stage. Finnur Jónsson (1905) argued against Storm on all three points. He showed convincingly that Storm had greatly exaggerated the closeness of Krm’s narrative content to that of Saxo’s account, and, while claiming an exclusively Icelandic origin for the poem, noted that the loss of <h> before <l> and <r> became in the twelfth century as much a feature of Old Norwegian as of Old Danish, and that occasional spellings reflecting it in Icelandic texts should cause no surprise. He further called into question Storm’s insistence on the lateness of Krm’s vocabulary, noting, for example, that the word bengrefill ‘wound-digger’ (st. 10/8), thought by Storm to be a German loan word, occurs in Egill Hfl 8/3V (Eg 41).
Finnur concluded that Krm was composed in Iceland c. 1200. De Vries (1964-7, II, 37-41), noting among other things a large number of verbal parallels (listed below, with additions) between Krm and RvHbreiðm HlIII, most likely composed by the Orkney jarl Rǫgnvaldr Kali Kolsson (d. 1158/59) and the Icelander Hallr Þórarinsson, maintained that Krm was composed in the Orkneys towards the end of the twelfth century. The parallels certainly suggest a close relationship between the two poems. Furthermore, the Maeshowe runic inscription number 23, which mentions the sons of Loðbrók, seems to indicate that part, at least, of the subject matter of Krm was known in Orkney in the mid-twelfth century (Barnes 1994, 178-86). On the other hand, as Olsen (1932a, 153; 1935, 78) notes, Rǫgnvaldr’s relatively traditional poetry, like that of his fellow Orcadian Bjarni Kolbeinsson (d. 1222), shows no sign of the lack of <h> before <l> and <r> which in Krm 29/7-8, for example, enables the alliteration of (h)læjandi ‘laughing’ with lífs ‘of life’, even though the relevant sound change took place in both the Orkneys and the Hebrides in the course of the twelfth century. The inference is that Krm was composed in an environment not far removed, but different, from that of Rǫgnvaldr Kali, and this is consistent with Olsen’s contention, accepted here, that it was composed in the Hebrides.
Structure, style and metre
Krm as a whole is virtually unique in skaldic poetry in having an initial refrain in all but one (st. 29) of its stanzas; another example is Ásb Ævkv (OStór), which has an initial refrain in seven of its nine stanzas. (On the absence of the refrain from st. 29 of Krm, see that stanza, Note to [All].) As in Ásb Ævkv, the refrain in Krm participates in the poem’s metrical scheme, forming a couplet (by alliteration) with l. 2 of each stanza in which it occurs (cf. Jesch 1998a, 215); in Krm the alliteration is on <h>. Krm is also unique in having ten lines per stanza in all stanzas apart from sts 23 and 29, each of which has eight lines (see Notes to [All] to sts 23 (second Note to [All]) and 29). Ten-line stanzas are not themselves unique in skaldic poetry: it is arguable that each of the two stanzas RvHbreiðm Hl 9-10III, which are poorly preserved, consists of ten lines (see Note to [All] to each of those two stanzas), but there is no sign that they contain a refrain; and Hl 43 and 44 may be said to consist of ten lines in that each takes the form of a normal dróttkvætt stanza with a refrain added as a fifth line at the end of each of its two halves: this refrain has internal alliteration and makes end-rhyme with the line preceding it. (See further Hl 43III, Note to [All].) The stanza’s structure is thus altogether different from that of the ten-line stanza of Krm.
The internal syntactic structure of the ten-line stanzas of Krm is clear-cut: Line 1 is the stef (except in the last stanza) and is line-stopped. Lines 2-4 belong together syntactically and semantically, and in most cases end with a period after l. 4 (with just two exceptions). Lines 5-8 follow the same pattern as ll. 2-4. Lines 9-10 are always a separate couplet syntactically.
Finnur Jónsson’s (1905, 160) view that Krm is closely related to the drápa form is questionable, at least as far as structure is concerned. Whereas Krm divides naturally into two main parts (sts 1-21, 22-9), with the same refrain running through both parts and dispensed with only in the final stanza of the second, the typical structure of a drápa (cf. Jón Helgason 1953, 110) is an introduction consisting of stanzas without a refrain; a middle section showing repetition of one or more refrains at regular intervals; and a concluding section which, like the introduction, has no refrain, and which may (but need not) be of the same length as the introduction. This again is very different from Krm. The poetic genre to which Krm is most readily assigned is that of the Sterbelied ‘death-song’ (Marold 2005b, 604-5), also known as ævikviða ‘autobiographical poem’ (Naumann 2001, 300). The latter Modern Icelandic term was chosen by Finnur Jónsson (in preference to ævidrápa) as appropriate for ǪrvOdd Ævdr (Ǫrv 71-141), one of the five poems, including Krm, considered by him (as also by Marold 2005b) as examples of death-songs (see LH 1894-1901, II, 149; LH II, 147-55). The second element in the compound ævikviða, with its connotations of the kviðuháttr metre and of eddic poems in fornyrðislag (such as Þrymskviða) does not, apparently, disqualify either Ásb Ævkv (OStór) (composed partly in dróttkvætt) or Krm (composed in a variant of dróttkvætt) from membership of the genre, cf. Konráð Gíslason (1881, 188-9). One may recall here the alternative title Loðbrókarkviða that Krm is given in 6ˣ, see above.
Krm is also anomalous in making sporadic and inconsistent use of hendingar, though their use is regular in the final couplets of nine of its twenty-nine stanzas, sts 1-3 (see Note to st. 3/9-10), 4, 7, 12, 15-17. By ‘regular’ is meant with skothending in the odd-numbered lines, aðalhending in the even-numbered. On the debatable cases of sts 5/9-10 and 10/9-10, see Notes to sts 5/9 and 10/9. On the relatively rare occasions that aðalhending occurs elsewhere in the poem, it occurs, as would be expected, in even-numbered lines: sts 6/10; 8/2, 10; 9/6; 19/6; 21/10; 23/4; 28/4; see, however, Note to st. 19/6, 7. Skothending occurs in odd-numbered lines (as would be expected) in the nine regular final couplets listed above, and also in the refrain that forms the first line of each of sts 1-28, but occurrences of skothending in odd-numbered lines are otherwise relatively rare: 4/7; 11/9; 13/7; 16/3, 7; 21/5; 28/5. Very frequent, on the other hand, are irregular occurrences of skothending in even-numbered lines; of these there are over thirty clear cases: sts 2/6, 8; 3/2, 4, 6; 4/2; 6/4, 6; 7/2, 6, 8; 9/8; 10/6, 8; 12/4; 13/2, 8; 15/6; 16/4; 17/4; 18/4, 8, 10; 20/6; 21/8; 22/4; 23/8; 24/2, 8, 10; 26/8, 10; 27/6, 10; 29/8.
Frequent also in Krm are instances of first person plural forms (whether of verbs or pronouns) seeming to have singular reference (to the speaker of the poem); it is possible, for example, that they do so in the refrain. Is this to be understood as ‘we hewed with the sword’ or ‘I hewed …’? While it is obvious in some cases (sts 1/3-5, 3/3, 26/6) that the reference is to the speaker of the poem and no one else, this is by no means always the case. In view of this uncertainty, the present edition follows the example of Jesch (1998a), in translating plural as plural and singular as singular in cases of first person forms of verbs, pronouns and possessive adjectives formed from pronouns, leaving it to the reader to decide whether the references are to the speaker alone, or to the speaker in company.
Questions of borrowing and imitation
Krm is particularly rich in verbal and phrasal parallels with other poems in the skaldic corpus, most especially with RvHbreiðm HlIII (see above). As indicated at appropriate points in the Notes, de Vries (1938, 722 n. 78; 1964-7, II, 40 n. 68) has noted parallels with this poem in Krm 1/10, 4/10, 7/7, 7/10, 9/3, 11/7, 16/10, 17/6 and 18/2-3. To his lists of parallels with RvHbreiðm HlIII may be added Krm 8/10, 9/6, 12/7, 16/7, 24/6 and 26/10, also pointed out in the Notes. The Notes also take account of parallels noted by de Vries between Krm on the one hand and, on the other, Bragi ÞórrIII (in Krm 11/3); Egill Hfl 8V (Eg 41) (in Krm 10/8), several of Egill’s lausavísur (in Krm 6/7, 23/3 and 25/6); Anon DarrV (Nj 53-63) (in Krm 6/3); Eskál LvI (by emendation, in Krm 5/5); Hfr HákdrIII (in Krm 12/6, 8), Hfr ÓldrI (in Krm 16/3-4), Hfr ErfÓlI (in Krm 12/8, 15/7), and Hfr Lv 10V (Hallfr 13) (in Krm 24/4); Sigv AustvI (by emendation, in Krm 5/5), Sigv BervII and/or Sigv ErfÓlI (in Krm 28/3); ESk Hardr III and/or ESk GeislVII (in Krm 22/6, 8), ESk GeislVII (in Krm 19/6) and ESk ElfvII (in Krm 4/7-8). To this list may be added the cases of Þorm Lv 25/5I (paralleled in Krm 9/6); ÞjóðA SexII (in Krm 3/7-8); Ragn 26 (in Krm 24/5-6); OStór 11/3 (in Krm 23/3); and Ketilr Lv1/3V(Vígl 2) (in Krm 20/4), also signalled in the Notes.
De Vries (1938) has further noted relatively few parallels of wording between Krm and poems in the eddic corpus, as follows: Akv (in Krm 21/3); Grott and Guðr I (in Krm 20/4); Ghv (in Krm 2/10). To this list may be added: Akv and HHund II (in Krm 26/4); Sigrdr (in Krm 7/9). All these instances are signalled in the Notes below.
In a category of their own, signalled but not specified in the Notes below, are parallels in skaldic poetry to those conventional passages in Krm (13/9-10, 18/5-6 and 20/5-6, 9-10) in which the hardships of warlike activity are contrasted with the pleasures of women’s company. These occur in KormǪ Lv 36V (Korm 55), Sigv Nesv 7I, and Anon Bjark 2/5-8III, as de Vries (1964-7, II, 38 n. 61) notes; he had earlier noted (de Vries 1927a, 57-8 and 58 n. 2) the parallel cases of Vígf LvI, Tindr Hákdr 1I, Rv Lv 17II, and KormǪ Lv 61/1-4V (Korm 82). Kock (NN §1278) also notes, in relation to Krm 20/5-6, the cases of Tindr Hákdr 1I and KormǪ Lv 61V (Korm 82), and at NN §343 points out further parallels in Heiðr 105, Þklypp Lv 1I, and Bbreiðv Lv 2/1-6V (Eb 25).
Olsen (1933a) has argued for the influence on Krm of Anon DarrV (Nj 53-63) and Sigv NesvI (both included in the lists above). Arguing that the battle referred to in Krm 11/5-7 took place on a Sunday, and noting that in Krm 11/7, 15/10 and 18/7 kennings occur in which the base-words mean ‘divine service’, ‘palm-tree’ and ‘ass’ respectively (see the Notes to those three stanzas), he maintains that these lines hint at the Christian feast of Palm Sunday, the commemoration of Christ’s entry into Jerusalem on an ass. They thus hark back, in Olsen’s view, to the Palm Sunday context of Sigv NesvI, in which S. Óláfr’s victory at Nesjar in Vestfold on Palm Sunday, 26 March 1016, is celebrated, and where reference is made both to Palm Sunday and to Easter in the final stanza (Sigv Nesv 15/2-4I). While it is conceivable that the poet of Krm had Sigv NesvI in mind when composing sts 11, 15, and 18, Olsen’s argument that the lines also reflect the Good Friday context (as he sees it) of Anon DarrV is much more doubtful. This poem may indeed have influenced Krm (cf. the Note to Krm 6/3), but not necessarily in the way Olsen is here suggesting: it makes no mention of any Christian feast, and the indications in Nj (ÍF 12, 454, 459) that it was inspired by the battle of Clontarf, fought on Good Friday 1014, are open to question (see the Introduction to Anon DarrV (Nj)).
In cases of similarity of wording between Old Norse poems it is uncertain whether one can speak of influence from one specific poem on another, or whether the similarities are due to a common stock of poetic diction on which poets could draw without having any particular previous poem in mind. Each case of similarity needs to be examined closely if it is to be used as evidence for the influence of one specific poem on another. Such influence, if it can be identified, may be helpful in dating poems, or at least in placing them in a chronological sequence. It is clear from de Vries’s remarks (1938, 722-3; 1964-7, II, 37-41; cf. 1927a, 57-8, 58 n. 2) that he would see the great majority of the parallels he lists (and no doubt also those added here to his lists of parallels with RvHbreiðm HlIII) as examples of Krm’s specific indebtedness to each of the poems he cites, showing Krm to be the borrower rather than the lender in each case. If he is right, at least in the cases of RvHbreiðm HlIIIand ESk ElfvII, then Krm must have been composed at the earliest in the second half of the twelfth century, since the former of those two poems dates from c. 1150 (Kuhn 1983, 317), and the latter from after 1159 (see SkP II, 565-6). The passage paralleling Krm in ESk Elfv 1/5-6II, may now be quoted: Elfr varð unda gjalfri | eitrkǫld roðin heitu ‘The bitter-cold Götaälv was reddened by the hot surge of wounds [BLOOD]’, and so may the corresponding couplet in Krm 4/7-8: öll var unda gjálfri | á sú roðin heitu ‘that whole river was reddened by a hot surge of wounds [BLOOD]’. Close reading of these two couplets in their contexts suggests that this is a case of one poem influencing another, and moreover that ESk ElfvII is the lender and Krm the borrower. The Elfv couplet forms an essential part of the information given in the eight-line stanza in which it occurs, establishing with its mention of the Götaälv the location of the battle that is being described, whereas in Krm the location of the battle – admittedly a more obscure one, the Ífa river – has already been established, in l. 5 of the ten-line stanza, where the prepositional phrase upp í Ívu ‘up the Ífa ’ is enough to indicate that a river is in question (see Fritzner: upp 2). The two lines in Krm are not essential to the overall meaning of the stanza in the way that those in ESk ElfvII are: in Krm they give the impression of being included for the purpose of filling out the stanza with the required number of lines. If Krm is borrowing in st. 4/7-8 from ESk ElfvII, a relatively young poem among those listed above as showing parallels to Krm, it may be tentatively suggested that the composition of ESk ElfvII, some time after 1159, provides a terminus post quem for Krm’s composition.
Editions and translations of Krákumál
The publication in Amsterdam of Ole Worm’s edition of Krm (in LR) in 1636, followed by a second edition in Copenhagen (Worm 1651, 182-207), unchanged apart from some additional misprints in the text of Krm and a few extra notes (cf. Rafn 1826, 75), profoundly influenced early editions and translations of the poem. The Latin translation in Worm’s edition, based on Magnús Ólafsson’s Danish translation, is by no means always accurate. This is especially apparent in all but the first of the six instances in Krm of varat ‘(it) was not’, i.e. sts 9/8, 13/9, 14/5, 18/5, 20/5 and 20/9, where ‑at is not recognised as a negative suffix. The result of this is that, in sts 13/9-10, 18/5-6, 20/5-6 and 20/9-10, warlike activity is presented in the translation not as a hardship to be contrasted with the company of women, but as a pleasure to be compared with it: fighting in battle is presented as just as pleasurable as bedding, kissing, or being served wine by a woman (it should be noted that Worm and subsequent early translators of Krm read as st. 14/9-10 what is printed in the present edition as st. 20/9-10, and print sts 19 and 20 in reverse order). This error is reflected in the work of early translators, including Lambert ten Kate’s (1723, I, 92, 93, 97) translation into Dutch, and (in part) Björner’s (1737, 48, 49) into Swedish. Mallet’s (1756, 152-4) translation into French is selective, and reproduces only the case of st. 20/9-10 (14/9-10 in LR), where the negative is missed. Percy in his English translation of 1763 (Clunies Ross 2001a, 114, 118, 120) skirts over the relevant lines in the case of st. 20/5-6, but clearly misses the negative in the three other instances. Sandvig (1779, 43, 44, 46) in his Danish translation misses all four of them, while Gräter (1789, 12, 15, 16), in his German translation fails to notice the negative in the case of st. 13/9, but in the other three cases, interestingly, seems to recognise that a negative is present, but translates it as a negative question expecting an affirmative answer (Wars nicht, als …? War nicht …? ‘Wasn’t it just as if …?’). All six of these translators further follow Worm’s text in mistranslating ór bjúgviðum hausa ‘out of the curved trees of skulls [DRINKING HORNS]’ in st. 25/6, taking it to mean ‘out of skulls’, thus conveying the preposterous idea of heroes in Valhǫll drinking out of the skulls of their enemies. All six of them convey correctly, however, as Worm does, the meaning of st. 29/8 læjandi skal ek deyja, as Percy’s translation (Clunies Ross 2001a, 130), ‘I die laughing’, illustrates. The far-reaching influence of these and other translations and mistranslations on the Romantic view of Germanic heroism in various European countries has been traced by, among others, Beck (1931), Gordon (1957, lxix-lxx), Terpstra (1960), Boyer (1972), Heinrichs (1978), Shippey (1998), Clunies Ross (1998c, 172-9, 231-6; 2001a, 90-131), McTurk (2007b), and Rix (2009, 55‑6).
The editio princeps of Krm is LR (Worm 1636), which is used in the present edition. Rafn (1826, 75-83), in his edition of the poem, which is also used in the present edition, lists twenty-eight publications containing text and/or translation of all or part of Krm. These include Worm (1636; 1651) and ten Kate (1723, I, 79-108). Rafn further lists Björner (1737, 43-51), which prints the poem, apparently transliterated from the text in Worm (1651; see Rafn 1826, 86), within the text of Ragn, with a Swedish translation to its right and a Latin translation, different from the one in Worm’s two editions, at the foot of the page. In addition to the translations by Mallet (1756), Percy (1763), Sandvig (1779) and Gräter (1789), Rafn’s list includes the edition and English translation of James Johnstone (1782), published together with a Latin translation and glossary into Latin and acknowledging (p. 2) the help of an anonymous ‘very learned native of Iceland’, confidently stated by Rafn (1826, 79) to be Grímur Jónsson Thorkelín (1752-1829); on this identification, see further Clunies Ross (1998c, 172-9). Rafn’s own edition (1826), which includes Danish and Latin translations of the poem, as well as a French translation by L. S. Borring, is the fullest and most responsible of early editions of the poem. It shows no awareness, it is true, of mss 147, R702ˣ, or R693, but gives pride of place among those it has used to 1824b and 6ˣ; it also takes LR into account (Rafn 1826, 83-8). It certainly has some of the limitations of its time but nonetheless contains many valuable observations and is referred to not infrequently in the Notes below.
In the present edition, LR (Worm 1636) is listed after each stanza together with the five mss used (see above), since readings are given from it as well as from them. The other editions used here, in addition to Skj, Skald, and Rafn (1826) (and listed with them as Editions after each stanza), have been found helpful for the following reasons: Pfeiffer (1860), for its glossary into German; CPB II, 339-45, for its introduction and English translation; Wisén (1886-9), for its glossary into Latin; Valdimar Ásmundarson’s edition (Krm 1891, 225-8) for its incorporation in the text of Krm of readings suggested by Jón Þorkelsson (1822-1904) and listed in the preface to Valdimar’s first edition (1885-9, I, 11); Finnur Jónsson (1893b), for its revised (rettet ‘corrected’) version of Wisén’s text, and for its notes to the text on p. 165; Finnur Jónsson (1905), for its accompanying essay on Krm and frequent textual divergences from Finnur Jónsson (1893b), as well as for occasional readings apparently abandoned in Skj B; and Olsen (Ragn 1906-8), which, although superseded as far as Krm is concerned by the present edition, is still of value for its diplomatic edition (Ragn 1906-8, lxxxiii-xcv, 176-94) of what Olsen could read of the 147 text of Ragn, of which Krm forms part.
All the editions and published translations of Krm mentioned above, with the exception of ten Kate (1723, I) and Mallet (1756), are listed, along with others, in Halldór Hermannsson (1912, 36-9; 1937, 61-2). Mention may additionally be made of the English translation by Schlauch (1930, xxxviii-ix, 259-67), listed in Halldór Hermannsson (1937, 62), and of the more recent translations by Jesch (1998a) and Waggoner (2009, 75-83) (into English), and by Renaud (2005, 91-106) (into French).