Ǫrvar-Odds saga (Ǫrv)
Skaldic vol. 8; ed. Margaret Clunies Ross
In total there are 141 extant stanzas or fragments of stanzas found in the various mss of Ǫrvar-Odds saga ‘The Saga of Arrow-Oddr’ (Ǫrv), an Icelandic fornaldarsaga that exists in three major versions of differing lengths. The mss also differ considerably in the number and arrangement of stanzas they contain. The disposition of the stanzas between versions is discussed below together with the implications of that distribution for the dating and authenticity of the poetry, which ranges across a variety of verse genres.
It has been generally accepted, following the two editions of R. C. Boer (Ǫrv 1888 and Ǫrv 1892), that the saga probably came into being in either the late thirteenth or the early fourteenth century (Ǫrv 1888, xxxvii), although a number of the extant ms. witnesses include material that is likely to be of later date, and this goes for the verse as well as the prose. Following the general policy of this volume of SkP, all the poetry from Ǫrv is assigned to the period 1250-1300, though some of it may be a lot older and some considerably younger. Indications of relative age are commented on in the Introductions to groups of stanzas and in Notes to individual stanzas as well as in Section 6 of the Introduction.
Ǫrv is a picaresque narrative about a legendary hero, Ǫrvar-Oddr ‘Arrow-Oddr’, who engages in a number of expeditions to exotic parts of the world, where he undertakes adventures and performs heroic deeds. In the surviving versions of his saga, Oddr is the direct descendant in the male line of Grímr loðinkinni ‘Hairy-cheek’, son of the legendary Ketill hœngr ‘Salmon’, about both of whom sagas were also composed (see Introductions to the poetry from Án, GrL and Ket in this volume). Oddr is thus represented as a kinsman of the men of Hrafnista (Ramsta), a small island near Nærøy in Folla off the coast of the northern Norwegian district of Nord-Trøndelag (see Þul Eyja 1/5III and Note). Common to the three sagas, GrL, Ket and Ǫrv, is the theme of the heroes’ travel from Norway to lands further north, whether in Finnmǫrk (Finnmark), where they engage with Saami people (Grímr, Ketill, Oddr) or, in Oddr’s case, in Bjarmaland (Permia), bordering the White Sea, where he has a number of memorable encounters with the local inhabitants. Boer (1892b, 98-106) argued that the saga of Oddr did not originally belong to the Hrafnista group (Án, GrL, Ket and Ǫrv), to judge by disparities between various aspects of the plot of Ǫrv and other sagas of the group. Oddr’s nickname Ǫrvar-Oddr ‘Arrow-Oddr’, is, however, linked to the Hrafnista legend, by means of a reference to Ketill hœngr’s acquisition of the three famous arrows called Gusisnautar ‘Gusir’s gifts’, from Gusir, King of the Saami, which Oddr’s father, Grímr, passes on to him (Ǫrv 1888, 24-5). In the late medieval ms. tradition, Ǫrv was clearly associated with other sagas of the Hrafnista group, and appears with them in the fifteenth-century ms. compilations 343a, 471 and the later paper copy 173ˣ (see below).
There are several key differences between Ǫrv and the other sagas of the Hrafnista group. The plot of Ǫrv is motivated by a sibyl’s prophecy that Oddr will live for three hundred years but be killed by the skull of his own horse on the farm of his foster-father in Norway, and that he will be doomed to wander from one place to another during the course of his long life. This prophecy gives narrators of the saga’s various versions the opportunity to embellish it with multiple adventures in distant lands and to introduce certain themes and characterisations of the hero that drew on models that may have been available from the biographies of other unnaturally long-lived figures, like the Wīdsīð of the Old English poem of that name (cf. Schlauch 1931) or the Norna-Gestr of Norn.
Another important ingredient in the genesis of Ǫrv is the episode of the battle on the Danish island of Sámsey (ModDan. Samsø) between a group of twelve berserk brothers and the hero and a single companion. The companion is killed in the fighting, while the hero, Oddr, survives. There are two significant analogues to this episode outside Ǫrv. The first is in Book V of Saxo Grammaticus’s Gesta Danorum, in which one Arvaroddus and his companion Hialmerus fight on Samsø against the twelve sons of Arngrimus (Saxo 2015, I, v. 13. 1-4, pp. 340-5) and the second comes in the incorporation of the episode into the narrative of Heiðr, another fornaldarsaga where the battle is motivated by rivalry over a woman, a motif absent from Ǫrv. Both these analogues, as well as Ǫrv itself, attest to the existence of a version of a narrative of the battle on Samsø involving Oddr, incorporating both verse and prose.
The earliest ms. of Ǫrv is Holm perg 7 4° (7), where Ǫrv is found on fols 43v-57r. Boer (Ǫrv 1888, i) adopts the siglum S for this ms. Ms. 7 is an Icelandic ms., usually dated to the beginning of the fourteenth century, which also contains texts of Konr (1r-9v), HG (10r-27v), Jvs (27v-39r), Ásm (39r-43r) and a fragment of Eg (57r-58v). The version of Ǫrv presented in 7 is considerably shorter than that in other mss, and it lacks certain sections of the narrative, notably the account of the battle on Samsø, although, as Boer (Ǫrv 1888, i) argued, there is palaeographical evidence to suggest that this episode was in 7’s exemplar, which Boer (Ǫrv 1888, xxxiv) maintained was a copy of the original saga text. In spite of its lacunae, 7 has been used as the base text for Boer’s 1892a edition and for one of two parallel texts, the other being 344a (see below), in his edition of 1888. The version of Ǫrv in this ms. contains forty-two stanzas or fragments of stanzas.
High on Boer’s stemma, but from a branch he called z, parallel to, but not descended from 7, is AM 344 a 4° (344a), of c. 1350-1400, probably written in Iceland, and originally combined with a text of Alex (now AM 519 a 4°). Árni Magnússon separated the two after he had obtained the ms. in Norway in 1689 (Ǫrv 1888, ii; Kålund 1888-94, I, 579 and 673). Boer referred to 344a as M and used it as one of two parallel texts in his 1888 edition. This ms. occupies an intermediate position on the stemma between 7 and a group of fifteenth-century mss that present a different and longer version of Ǫrv. Ms. 344a includes sixty-four stanzas or parts of stanzas and is divided into three major sections (þættir) and, within these, into chapters, many introduced with scribal headings and coloured initials.
Below 344a on Boer’s stemma is a group of mss arguably derived from an exemplar different from, but parallel with it, which he designates a. The fifteenth-century vellum fragment AM 567 IV 4° (567IV) contains a segment of GrL on fol 1r-1v and a passage from Ǫrv on fols 2r-3v, including three stanzas on fol 2v. Below 567IV on the stemma, but, according to Boer, derived from lost intermediaries parallel with it, are the fifteenth-century mss AM 343 a 4° (343a), designated A by Boer, AM 471 4° (471), designated B by Boer, and AM 173 folˣ (173ˣ), designated E by Boer, the last, a paper ms. of c. 1700, related to the other two but deriving from a different intermediary exemplar. It is these three late mss, 343a, 471 and 173ˣ, that contain the greatest number of stanzas associated with Ǫrv, and this tendency towards amplification is most obvious in Oddr’s so-called Ævidrápa ‘Life drápa’, more correctly an ævikviða ‘life poem’, as it has no refrain. Ms. 343a is an anthology of sagas, most of them either fornaldarsögur or riddarasögur, while 471 contains seven saga texts, three late sagas of Icelanders (Þórð, Krók and Kjaln), three sagas of Hrafnistumenn (GrL, Ket, and Ǫrv) and one riddarasaga, Vikt. Ms. 173ˣ contains Ket, GrL, Ǫrv, Án and Frið.
Many questions arise from a consideration of the identity and disposition of stanzas associated with Ǫrv in the medieval mss in which the saga is recorded and later, paper copies, derived from them. From a consideration of the ms. witnesses, we can be certain that some version of the exchange between Ǫrvar-Oddr and a prophetess named Heiðr existed in the earliest version of the saga, represented by 7, and it is probable that stanzas associated with the narrative of the battle on Samsø were also included in early versions. Although 7 omits this section of the saga, 344a has all of the stanzas relating to this narrative, and a good many of them, often in significantly different versions, are also attested in two mss of Heiðr, GKS 2845 4° (2845) and UppsUB R715ˣ (715ˣ) (Heiðr 1924, 9-14, 97-101; Heiðr 1960, 5-9, 69-75). Heiðr is generally thought to have been composed in the first half of the thirteenth century. As we have seen, there is also an analogue to the battle on Samsø in the legendary material Saxo Grammaticus used in Book V of his Danish history. The kernel of the Samsø narrative and accompanying stanzas is likely to be old, but how old it is and the extent to which it was originally associated with the character of Ǫrvar-Oddr, is difficult to determine; on this see Boer (1892b, 112-23; cf. Heiðr 1924, lxvii-lxix). Certainly, the extent of variation between the stanzas that have been preserved about this battle and its aftermath point to the influence of an active oral tradition.
Another sequence of stanzas which has a good claim to have existed in some form in early versions of Ǫrv is the mannjafnaðr ‘comparison of men’ that accompanies a drinking contest between Oddr and two of the retainers of a certain King Herrauðr (cf. Boer 1892b, 127-34). The version of 7 includes twenty-five stanzas belonging to this sequence. On the other hand, 7 has only eleven of the stanzas that have traditionally been included in Ǫrvar-Oddr’s seventy-one-stanza Ævidrápa, which he is supposed to have composed at the end of his life, and these stanzas (ǪrvOdd Ævdr 21, 22, 38, 41, 53, 52, 64, 69, 68, 70, 71) occur at different points in the sequence of the narrative, in conjunction with the events they describe. The first four (21, 22, 38 and 41) also occur at comparable points in the other mss. Boer (Ǫrv 1888, xv) argued that these stanzas may originally have been lausavísur, which were later pressed into the service of compilers’ desires to fill out the Ævidrápa with more and more examples of high points in Oddr’s career. The continuous Ævidrápa occurs only in 343a and 471, and the major part of it is also in 173ˣ. This poem is likely to have grown by accretion (though Boer attributes the entry of lausavísur that formed the nucleus of the Ævidrápa to the redactor of z). It is possible, as Boer (Ǫrv 1888, xvii) suggests, that the dying Oddr was originally credited with a short poem, of which twelve lines survive in 7 (Ǫrv 140/5-8 and 141), and that this was then greatly elaborated. Boer (Ǫrv 1888, xv; Boer 1892b, 125-7) labelled the Ævidrápa, as it appears in 343a, 471 and 173ˣ, an interpolation, but it is probably more helpful to our understanding of changing tastes in poetry in Iceland during the fourteenth and fifteenth centuries to term it an amplification of a nucleus of stanzas that were present in earlier versions of the saga or in poetry associated with it.
The stanzas in Ǫrv have been edited a number of times, both as part of editions of the whole saga and separately. The editio princeps of Ǫrv was that of Olaus Rudbeck (1697), with a Latin translation. It was then edited by Rasmus Rask (1819) and by Rafn in FSN 2, 161-322 and 504-59 (the latter a reprint of Rask’s text). There is also a text of the version of Ǫrv based on the younger mss in FSGJ 2, 199-363. The standard modern edition of Ǫrv is Boer’s Ǫrv 1888, which gives the parallel texts of 344a and 7, with some additional episodes from 343a. He also produced an edition, using 7 as his base text, but relying on other mss where necessary, for the Altnordische Saga-Bibliothek (Ǫrv 1892). Christopher Tolkien gives an edition of the Samsø stanzas from 344a’s version of Ǫrv in Appendix A to his 1960 edition of Heiðr (Heiðr 1960, 69-75). A. LeRoy Andrews undertook a series of significant studies between 1914 and 1927 of the mss of Heiðr and made some valuable observations on these and on their relation to Ǫrv and the stanzas they have in common. In the present edition of the stanzas from Ǫrv, only Ǫrv 1888, Ǫrv 1892, and FSGJ are cited routinely in the Editions line. Where the stanzas also occur in Heiðr, the editions of Jón Helgason (Heiðr 1924) and Christopher Tolkien (Heiðr 1960) are also cited. Also listed in the Editions line are the separate editions of the stanzas only in Edd. Min., Skj A and B, Skald and Kock’s commentary in NN. CPB I, 159-63 has a text of Hjálmarr’s death-song (see Introduction to Ǫrv 13-29 below), but this is not used in the present edition.
Previous editors have divided and arranged the stanzas from the different versions of Ǫrv in ways that make it difficult for the reader to understand how they lie in the mss themselves. Edd. Min. divided stanzas into groups according to Heusler’s and Ranisch’s perception of their cohesion as lieder ‘songs’, the remnants of ancient poems, but not primarily according to the overall sequence of the stanzas in the mss themselves. Finnur Jónsson (Skj A and B), followed by Kock (Skald), adopted a different strategy. To some extent Finnur arranged the stanzas according to their sequence in Boer’s 1888 edition, which follows 7 and 344a, with page references to Ǫrv 1888 and Edd. Min. in the notes to Skj A. Finnur begins with Ǫrv 1-3 (although only stanzas 2-3 are in 7) (= Finnur’s sequence I), following this with Ǫlvǫr Lv 1 (Ǫrv 4) (sequence II). This is not in fact the next stanza to occur in 7, as three stanzas, found in 7 and the other mss as lausavísur, but also occurring as stanzas from Oddr’s Ævdr in the younger mss (viz. Ævdr 21, 22 and 38), come before Ǫlvǫr Lv 1. Furthermore, in the saga Oddr’s response to Ǫlvǫr, which is found as a lausavísa in 7 and all the other mss, also appears as Ævdr 41 in the younger mss. In Skj A and B, Ǫlvǫr Lv 1 is followed by Finnur’s sequences III and IV and two supplements (a and b), which comprise the stanzas associated with the fight on Samsø and Hjalmarr’s death-song. None of these stanzas appear in 7, though many are in 344a and even more in the younger mss. Some of them are also in the Heiðr mss 2845 of c. 1450 and R715ˣ of c. 1650, but are not edited separately by Finnur or Kock with other stanzas from Heiðr, only with those from Ǫrv. However, where the Samsø stanzas are found in both sagas, Finnur privileges the text of Heiðr in 2845 and makes no use at all of R715ˣ; for Ǫrv he uses 344a as his main ms. and gives supplementary stanzas from 343a and 471. Finnur’s sequence V comprises one helmingr, Ǫrv 31, which is only extant in an account of Oddr’s visit to Risaland (Giantland) in the younger mss (cf. Ǫrv 1888, 121), while his sequence VI includes three stanzas from another episode only in the younger mss. Meanwhile 7 has, as lausavísur, Ævdr 53 and 52 and, somewhat later, Ævdr 64. Finnur’s sequence VII comprises twenty-five stanzas associated with Oddr’s mannjafnaðr at the court of King Herrauðr. Here he uses 7 as the main ms. Finnur’s sequence VIII involves a hostile exchange between Oddr and a heathen priestess in Bjálkaland, which appears in one form or another in all mss. 7’s version of a helmingr that also appears as Ævdr 69 comes before the Bjálkaland episode in that ms., and its version of the stanza that is also Ævdr 68 comes at the episode’s end. Finnur’s final sequence IX comprises Oddr’s Ævdr, attested in extended form (seventy-one stanzas) in the younger mss. Here, as mentioned above, 343a is his base ms.
Because of the difference between versions of Ǫrv, the overlap with stanzas from Heiðr and the variation in the number, form and sequence of stanzas in each version, it is difficult to display their sequencing so that the reader can grasp where they come in the various mss, and how the Ævdr is likely to have grown from lausavísur connected with specific episodes in the narrative into an entire poem, as it appears in the younger mss. The three Tables below, attached to the Introductions to the Samsø episode, Hjálmarr’s Death-song and Oddr’s Ævdr respectively, set out the order in which the stanzas belonging to these sequences occur in the mss themselves and the distribution of stanzas within mss.
The great variation between the ms. witnesses to Ǫrv stanzas has necessitated fairly frequent changes in editorial decisions about the ms. to be chosen as base text for any given stanza. In general, however, 7 is taken as base ms. where it records a stanza, failing which 344a is usually chosen as base. However, in the stanzas that Ǫrv and Heiðr have in common, 2845 is sometimes given preference over 344a. Where stanzas are not recorded in either 7 or 344a, 343a is usually chosen as base ms., though sometimes, if its text is defective, 471 provides the base text.