Skaldic Poetry of the Scandinavian Middle Ages

login: password: stay logged in: help



Vol. I. Poetry for Scandinavian Rulers 1: From Mythological Times to c. 1035

3.1. Sagas of the kings of Norway to c. 1035

1. Ágrip af Nóregs konunga sǫgum (Ágr) (DW)
2. Fagrskinna (Fsk) (DW)
3. Flateyjarbók (Flat) (DW)
4. The Greatest Saga of Óláfr Tryggvason / Óláfs saga Tryggvasonar in mesta (ÓT) (DW)
5. Hauksbók (Hb) (DW)
6. Heimskringla (Hkr) (DW)
7. Hulda and Hrokkinskinna (H-Hr) (DW)
8. Jómsvíkinga saga (Jvs)
9. The Legendary Saga of S. Óláfr / Helgisaga Óláfs konungs Haraldssonar (ÓHLeg) (DW)
10. Morkinskinna (Mork) (KEG)
11. Óláfs saga Tryggvasonar by Oddr Snorrason (ÓTOdd) (DW)
12. The Oldest Saga of S. Óláfr / Óláfs saga helga in elzta (ÓHÆ)
13. The Separate Saga of S. Óláfr / Óláfs saga helga in sérstaka (ÓH) (DW)

(Vol. I. Poetry for Scandinavian Rulers 1: From Mythological Times to c. 1035 > 8. Volume Introduction > 3. Sources for skaldic poetry cited in the kings’ sagas: manuscripts, facsimiles and editions > 3.1. Sagas of the kings of Norway to c. 1035 > 6. Heimskringla (Hkr))

6. Heimskringla (Hkr) (DW)

Stemma (from ÍF 28, xciv)

Manuscripts: The Kringla group (x class)

K:        Kringla, almost destroyed in the Copenhagen fire in 1728. One leaf remains, Lbs frg 82, formerly Holm perg 9 I fol, in Kungliga biblioteket, Stockholm (Icelandic, c. 1258-64). Copies of Kringla made before the fire:

a) :     Comprising AM 35 folˣ = Hkr I (used in SkP I), AM 36 folˣ = ÓHHkr (Hkr II, used in SkP I), AM 63 folˣ = Hkr III (used in SkP II). All are c. 1675-1700, in the hand of Ásgeir Jónsson. Normally considered to give the most reliable text (Jørgensen 2000, 231-2; Jørgensen 2007, 318).

b) 521ˣ: OsloUB 521ˣ (c. 1700). Hkr I copied from K, Hkr II from Jöfraskinna (J), but from K where J had lacunae.

c) 70ˣ:    AM 70 folˣ (c. 1675-1700; Hkr II only)

d) papp18ˣ: Holm papp 18 folˣ (c. 1650-1700, in the hand of Jón Eggertsson).

e) J2ˣ:     AM 38 folˣ (c. 1675-1700; by Ásgeir Jónsson). K was used in places to fill out lacunae; see Jöfraskinna group and discussion below.

f) Bb:     Bergsbók, Holm perg 1 fol (Icelandic, c. 1400-25). Partly a Hkr ms: see discussion below.

39:       AM 39 fol (Icelandic, c. 1300). 43 leaves. The first ten leaves have defective and difficult text from the first part of Hkr. The remainder extends from ÓHHkr ch. 245 to HsonaHkr ch. 2.

22ˣmarg: Holm papp 22 folˣ, marginalia (early C17th). Variants in the margins of the ÓT ms. 22ˣ, apparently copied from a now lost folio of 39 (Ólafur Halldórsson 1976; ÓT 1958-2000, III, ccx).

F:         Codex Frisianus or Fríssbók, AM 45 fol (Icelandic with Norwegian features, c. 1300-25). Preserves Hkr I and III, but not Hkr II, i.e. ÓHHkr. Contaminated by a text from the y-branch.

Manuscripts: The Jöfraskinna group (y class)

J:         Jöfraskinna, almost destroyed in the fire in Copenhagen in 1728. Four leaves from Óláfs saga helga survive, Holm perg 9 II fol (Icelandic, c. 1300-25). Other fragments in AM 325 VIII 3 d 4° and NRA 55 A (both used in SkP II). Copies of J made before the fire:

a) J1ˣ:     AM 37 folˣ (made 1567-8 by Jens Nielsen, and continued c. 1700). Óláfs saga helga follows ÓH, not Hkr redaction; ends at ch. 74.

b) J2ˣ:    AM 38 folˣ (c. 1675-1700; by Ásgeir Jónsson). Óláfs saga helga follows the Separate ÓH redaction, not Hkr; lacunae filled from K ms. of Hkr.

E:        Eirspennill, AM 47 fol (c. 1300-25). Considered the best of this group. Contains the last chapter of ÓHHkr, Hkr III, Sverris saga and Hákonar saga Hákonarsonar. Used in SkP II and minimally in this volume.

325VIII 1: AM 325 VIII 1 4° (c. 1300-25). Fragment.

325XI 1: AM 325 XI 1 4° (c. 1300-25). Fragment.

G:        Gullinskinna. Destroyed in the Copenhagen fire of 1728. One leaf from c. 1400 preserved in AM 325 VIII 5 c 4° (325VIII 5 c). Used in SkP II. G was copied in:

42ˣ:      AM 42 folˣ (c. 1675-1700; in the hand of Ásgeir Jónsson). Used in SkP II.

325XI 2 e, 2 f: AM 325 XI 2 e (c. 1250-1300), AM 325 XI 2 f (c. 1350-1400). Fragments containing one stanza each edited in SkP II and tentatively assigned there to Hkr (see SkP II, lxxix).

Facsimiles and editions: K 1895, F 1932, J 1895; Hkr 1893-1901, I-IV, ÍF 26-8, Hkr 1991, I-III [Modern Icelandic orthography], F 1871, E 1916, Veland 1997, Yng 2000, Jørgensen 2007, Kyrkjebø 2001.

Hkr is a cycle of sagas about Norwegian kings from prehistorical times to c. 1177. It takes its (modern) title from the words heims kringla ‘The circle of the world’ with which the opening geographical description starts in the Kringla ms. The Icelandic chieftain Snorri Sturluson, author of SnE, is normally credited with Hkr’s composition c. 1230. Although no medieval source attributes Hkr to Snorri there is a good deal of evidence in favour of his authorship (summarised, e.g., in Whaley 1991, 13-19), and it is accepted throughout this volume, though the theory of later compilation noted below should be kept in mind.

When Snorri came to write Hkr the stories of the earlier kings of Norway had undergone a long process of textualisation: writing down, compiling, editing and refining. His sources were therefore primarily written. Although Snorri emphasises ‘model predecessors’ (Whaley 1999, 242) such as Eiríkr Oddsson’s *Hryggjarstikki (c. 1150) and Ari Þorgilsson’s Íslendingabók (before 1133), he in fact depended heavily on late twelfth-century or thirteenth-century works such as Ágr, Fsk, Fær, Mork, ÓHLeg, Orkn, ÓTOdd and Skjǫldunga saga (Skjǫld), often in earlier versions than those known to us. Also used were a now-lost *Hlaðajarla saga and the Lífssaga Óláfs helga written by Snorri’s contemporary Styrmir Kárason (on which, see Flat above). Hkr’s dramatic yet sober and plausible style and the well-integrated organisation of its material have been much admired, though recent scholarship emphasises how these qualities also frequently characterise Snorri’s sources. The high medieval estimation of Hkr is demonstrated both by the number of times it was copied, and by the indebtedness to it of later compilations such as the Greatest Saga of Óláfr Tryggvason (ÓT) and Flat.

The centrepiece of Hkr and its longest component saga is Óláfs saga helga (ÓHHkr), and Hkr falls naturally into three parts: I, from the Prologue to Óláfs saga Tryggvasonar, ending with the battle of Svǫlðr (c. 1000); II, Óláfs saga helga; III, from Magnúss saga góða to Magnúss saga Erlingssonar, i.e. from c. 1035 to the accession of Sverrir Sigurðarson in 1177. The comments here focus on parts I and II, as major sources for poetry in SkP I; for further details of mss of part III and the stanzas it contains, see SkP II, lxiii-lxv.

The mss that are relevant to the editions in SkP I are the following: (AM 35 folˣ, 36 folˣ, 63ˣ (minimally)), 39, F, J1ˣ, J2ˣ, E (minimally) and the fragmentary 325VIII 1, 325XI 1. The ms. preservation of the three parts of Hkr differs radically, and of the mss shown in the composite stemma above only contains all three parts. Nevertheless, for parts I and III the ms. relationships can be represented as a two-branched stemma, comprising the x or Kringla class, thought to be closer to the archetype, and the y or Jöfraskinna class. The exemplars of both classes are lost, except for a single vellum leaf of Kringla (K), which preserves a text of Ótt Knútdr 11 and ÞSjár Róðdr, and four leaves of Jöfraskinna (J), which preserve Ótt Hfl 18/8 and 19. Within the x class the K text is closest to the lost exemplar and is almost complete, and Ásgeir Jónsson’s careful copies in are the most reliable witness to K. The transcripts papp18ˣ and 521ˣ have also been consulted selectively in this edition, particularly where it is important to establish the reading of K (see Jørgensen 2000, 232 for a list of the transcripts of K covering Hkr I ranked according to accuracy). Text from K was also copied into J2ˣ to fill lacunae in both Hkr I (Hgráf ch. 13 to ÓTHkr ch. 63) and Hkr II (five smaller sections; see Jørgensen 2000, 39, 232). Further, the text of ÓHHkr chs 57-119 (ÓH chs 43-107) in the ÓH ms. Bergsbók (Bb) was evidently copied from a ms. close to K, and hence is a valuable witness in cases of disagreement between K and the ÓH mss (ÓH 1941, II, 1116). The Hkr I text in F descends from the same x-class exemplar as that of 39, but the scribe was also evidently influenced by a y-class text (Ólafur Halldórsson 2001, xxxi-xliii, lvi). Much of Hkr I is preserved only in the K transcripts, though the ÓT mss, which contain y-class text, are also invaluable witnesses (Ólafur Halldórsson 2001, lvi). In the y class, Jöfraskinna was the closest ms. to the lost exemplar, and again there are careful paper copies, in J1ˣ and J2ˣ, of which J2ˣ has been regarded as superior. The text of Óláfs saga helga (= Hkr II) in Jöfraskinna is not that of Hkr, but rather a shortened version of an A-class text of the Separate ÓH, except for the lacunae mentioned above. Ms. E is the only well-preserved medieval representative of the y-class, but is used little in this volume since it commences close to the end of Hkr II while G contains only part of Hkr III so is not relevant to the poetry in this volume. A number of fragmentary mss, from one to five leaves in length, also exist (see stemma and table) but most are not relevant to this volume.

As noted above, of the known medieval mss only K contained all three parts of Hkr, perhaps because scribes or their patrons already owned a copy of the Separate Saga of S. Óláfr (ÓH), so could leave out Óláfs saga helga when copying Hkr. Indeed it has been argued that Hkr is not a unitary composition by Snorri Sturluson but an assemblage that post-dates him (Louis-Jensen 1997; cf. Jørgensen 2000, 70 and Kyrkjebø 2001, 128-34, who regards Hkr as Hkr I and III together with a version of the saga of Óláfs saga helga). On this view the Óláfs saga helga in K may be regarded not as the Hkr version of the saga but simply as the K version, with a similar status to that of the J version of Óláfs saga helga (Louis-Jensen 1997, 239 et passim; Jørgensen 2000, 10, 63-70).

The status of Hkr within the konungasögur ‘kings’ sagas’ calls for a brief comment here. The priority given in Skj and elsewhere to the evidence of Hkr is justified by such factors as the quantity of poetry preserved there, and the range and quality of ms. witnesses. However, just as Hkr as a literary and historical work has perhaps been prized among the kings’ sagas at the expense of the others, it is possible to exaggerate its importance as a textual witness, and the editing of the poetry needs to evaluate the evidence of Hkr mss carefully in the light of groupings such as the Fsk mss and those of earlier versions of the sagas of Óláfr helgi and Óláfr Tryggvason.

The textual relationship of Hkr to the Separate Saga of S. Óláfr / Óláfs saga helga in sérstaka (ÓH) is also an important and difficult issue. Since the publication of Sigurður Nordal’s 1914 monograph Om Olaf den helliges saga, it has been generally accepted that ÓH was the work of Snorri Sturluson and was the basis for the same saga in Hkr (see Whaley 1991, 54-5 for summary of the case). Jöfraskinna, a Hkr ms. in its parts I and III, has a text of Óláfs saga helga belonging to the A class of ÓH, while Kringla has an adapted version. This has therefore been regarded as a distinctive ‘Hkr’ redaction, and it is referred to as ÓHHkr in this edition, though as noted above a case can be made for regarding it merely as a distinctive ‘Kringla’ redaction. In either case, the implication for text-critical purposes is that where there is disagreement among the ÓH mss, a reading supported by the K transcripts of Hkr is likely to be the original, and this holds good in the great majority of cases. So much is stated by Johnsen and Jón Helgason (ÓH 1941, II, 1093), who also describe Kringla as having i det hele en god og opprinnelig tekst ‘on the whole a good and original text’. However, is also clear that even the lost original was not error-free (ibid., 1123) and that there are exceptions where the Kringla text is altered, sometimes in a way that by chance concurs with certain ÓH mss (ibid., 1093). Meanwhile no individual ms. or class of mss within the ÓH stemma gives a consistently reliable witness, for Holm2, the oldest and generally the best, has unique errors (ibid., 1103). Hence for practical purposes an edition might reasonably proceed by adopting either or Holm2 as a main ms, always bearing in mind that if we seek to get back to Snorri’s original text, the whole transmission must be investigated in every case (ibid., 1123).

Important work by Jonna Louis-Jensen takes the analysis a step further by suggesting that Kringla (K) might be accommodated within the ÓH stemma in the following way (1997, 240):


Hkr makes very extensive use of skaldic poetry, citing some 380 stanzas in parts I and II, and it is our richest source for Old Norse encomiastic poetry prior to c. 1035, even if, like Fsk, it took many of its citations from written sources (see Fidjestøl 1982, 29-30 for a list). Most appear as authenticating quotations of single stanzas, but there are also some extended extracts from long poems. Hkr I contains the whole of Þjóð Yt and is the sole source for all but one stanza; it also cites substantial amounts of Þhorn Gldr, Þhorn Harkv, Glúmr Gráf, Gsind Hákdr, Eyv Hák, Eskál Vell, Hfr Óldr, ÞKolb Eirdr, Tindr Hákdr, Edáð Banddr, Hókr Eirfl and Hfr ErfÓl,  together with some lausavísur. Hkr II has virtually the same complement of poetry as ÓH including the bulk of Sigvatr Þórðarson’s work: see the listing below. On the poetry in Hkr III, see SkP II, lxiv-lxv.

© Skaldic Project Academic Body, unless otherwise noted. Database structure and interface developed by Tarrin Wills. All users of material on this database are reminded that its content may be either subject to copyright restrictions or is the property of the custodians of linked databases that have given permission for members of the skaldic project to use their material for research purposes. Those users who have been given access to as yet unpublished material are further reminded that they may not use, publish or otherwise manipulate such material except with the express permission of the individual editor of the material in question and the General Editor of the volume in which the material is to be published. Applications for permission to use such material should be made in the first instance to the General Editor of the volume in question. All information that appears in the published volumes has been thoroughly reviewed. If you believe some information here is incorrect please contact Tarrin Wills with full details.