Margaret Clunies Ross 2017, ‘The Icelandic fornaldarsaga’ in Margaret Clunies Ross (ed.), Poetry in fornaldarsögur. Skaldic Poetry of the Scandinavian Middle Ages 8. Turnhout: Brepols [check printed volume for citation].
The term fornaldarsaga (pl. fornaldarsögur) ‘saga of the ancient time’, or mythical-heroic saga, refers to a sub-genre of the medieval Icelandic saga. Depending on the classification one adopts, the fornaldarsaga sub-genre includes approximately twenty-five sagas (cf. Torfi H. Tulinius 2005, 448; Clunies Ross 2010a, 31; or about thirty-five, according to the Stories for All Time: the Icelandic fornaldarsögur database). Twenty-one of these are represented in this edition, although two of them are sometimes regarded as riddarasögur ‘sagas of knights’ or romances, namely Máguss saga jarls ‘The Saga of Jarl Mágus’ (Mág) and Þjalar-Jóns saga ‘The Saga of File-Jón’ (ÞJ). Each of these two sagas displays some characteristics of the fornaldarsaga, however, including a few stanzas of poetry, unlike the riddarasaga sub-genre, which normally contains no verse. Poetry is an important component of the Icelandic fornaldarsaga, being found in about eighty percent of the corpus, and in a proportion varying from a low of one to three stanzas per saga to a high of 141 in the later manuscripts of Ǫrvar-Odds saga ‘The Saga of Arrow-Oddr’ (Ǫrv).
The term fornaldarsaga is not of medieval origin but was created by the Danish philologist Carl Christian Rafn (1795-1864) in the early nineteenth century, when, in 1829-30, he published a collection of thirty-one texts in three volumes that he called Fornaldar Sögur Nordrlanda ‘Sagas of the Ancient Time of the Northern Lands’ (FSN). Most of the narratives Rafn brought together have a number of characteristics in common that differentiate them from other sub-groups of the Icelandic saga. Firstly they are set in the prehistoric period, the fornǫld ‘ancient time’, and before the settlement of Iceland (c. 870-930 AD). Their protagonists may sometimes be claimed as ancestors of Icelanders from the historical period. Second, the geographical settings of fornaldarsögur are either in mainland Scandinavia, usually either Norway or Sweden, or in some more exotic location, such as the place called Bjarmaland in several sagas, which is usually identified with Permia, an area on the Kola Peninsula in the vicinity of the White Sea in northwest Russia. Thirdly, the fornaldarsaga world admits of a greater number of paranormal beings and happenings than one usually finds in some other kinds of saga literature, although this distinction is by no means hard and fast. Fourthly, and importantly for this edition, the majority of the fornaldarsögur contain poetry, and most of it is in eddic metres (fornyrðislag, ljóðaháttr and málaháttr for the most part), although variants of dróttkvætt are used in some sagas, most notably in Ragnars saga loðbrókar ‘The Saga of Ragnarr Hairy-breeches’ (Ragn) and in the anonymous poem Krákumál ‘Speeches of the Crow’ (Anon Krm). The metres of fornaldarsaga poetry are discussed in Section 6 below.
Various attempts have been made to distinguish sub-categories within the fornaldarsaga, the best-known being that of Helga Reuschel (1933), who identified three sub-categories of fornaldarsaga: heroic sagas, viking sagas and sagas of adventure. Hermann Pálsson (1985b), on the other hand, identified only two categories: heroic legends and adventure tales. While it may not be possible or desirable to establish hard-and-fast distinctions between groups of fornaldarsögur (Lönnroth 2003), it can be seen that the majority of these sagas that incorporate a good deal of poetry are heroic sagas, which depend on legends that are recorded in many cases in other early Germanic literary traditions and in the Gesta Danorum ‘History of the Danes’ (c. 1200) of Saxo Grammaticus. Sagas of this type include Gautreks saga ‘The Saga of Gautrekr’ (Gautr), Hálfs saga ok Hálfsrekka ‘The Saga of Hálfr and Hálfr’s champions’ (Hálf), Hervarar saga ok Heiðreks ‘The Saga of Hervǫr and Heiðrekr’ (Heiðr), Hrólfs saga kraka ‘The Saga of Hrólfr Pole-ladder’ (Hrólf), Ragn and Vǫlsunga saga ‘The Saga of the Vǫlsungar’ (Vǫls).
In some of the poetry in this group of sagas clear thematic and verbal parallels can be established with comparable verse in other languages, as, for example, between the prose narrative and stanzas recorded in Ásmundar saga kappabana ‘The Saga of Ásmundr, Slayer of Champions’ (Ásm), the Old High German Hildebrandslied ‘Lay of Hildebrand’ and Book VII of Saxo’s Gesta Danorum. Likewise, incidents described in some heroic fornaldarsögur are attested indirectly by early Norse skaldic poetry, a case in point being the fire-kenning bani Hôalfs ‘slayer of Hálfr’ recorded in the late ninth- or early tenth-century poem Ynglingatal ‘Enumeration of the Ynglingar’ (Þjóð Yt 6/7I), in an allusion to the hall-burning of the legendary king Hálfr and his men, as told in Hálf.
On the other hand, sagas set in the Viking Age, like Friðþjófs saga ins frœkna ‘The Saga of Friðþjófr the Bold’ (Frið), and more fantastic adventures involving mistaken identity, shape-shifting, bridal quests and similar motifs, such as we find in Hjálmþés saga ok Ǫlvis ‘The Saga of Hjálmþér and Ǫlvir’ (HjǪ), Mág, and Sturlaugs saga starfsama ‘The Saga of Sturlaugr the Industrious’ (StSt), have no parallels in cognate early Germanic literatures or earlier Old Norse witnesses and are not likely to be of very great age.
The so-called Hrafnista group of fornaldarsögur (named for the Norwegian island from which the family of the Hrafnistumenn is said to have come), Áns saga bogsveigis ‘The Saga of Án Bow-bender’ (Án), Gríms saga loðinkinna ‘The Saga of Grímr Hairy-cheek’ (GrL), Ketils saga hœngs ‘The Saga of Ketill Salmon’ (Ket) and Ǫrvar-Odds saga ‘The Saga of Arrow-Oddr’ (Ǫrv), forms a specific unit within the fornaldarsaga corpus and owes its cohesion as a group both to internal genealogical links between the protagonists, to thematic and geographical parallels, and to external links claimed with Icelandic families of the settlement age. Ǫrv, while belonging within the Hrafnista group, also has characteristics of both the heroic group and the fantastic adventure type of saga.
Although the earliest extant manuscript witnesses to the fornaldarsaga genre date from the early fourteenth century (see Section 4 below), there are several kinds of evidence that support the idea that this sub-genre of the saga may have been among the first to develop in Iceland during the twelfth and early thirteenth centuries. This was a period of intense interest among Western European writers in constructing the legendary past of their own communities, as witness Geoffrey of Monmouth’s De gestis Britonum ‘Concerning the deeds of the Britons’ (c. 1136), clearly known to some Icelanders in the twelfth century (see Section 3 below), and the Gesta Danorum of Saxo Grammaticus, completed c. 1200 and, by Saxo’s own admission, based in part on ‘their packed store of historical treasures’ (quorum thesaurus historicarum rerum pignoribus refertos) that his Icelandic informants had made available to him (Saxo 2015, I, Pr. 1. 4, pp. 6-7). Indeed, although Saxo’s treatment of his sources is different from the treatment of similar subjects found in fornaldarsögur, the many parallels between them are significant. They seem to indicate that the combination of heroic poetry with a prose narrative frame had become a viable means for the literary expression of the subject of legendary prehistory, whether in Latin or a vernacular language, by about 1200 (cf. Torfi H. Tulinius 2002, 48-69). Around the same time sagas with subjects from the historical period after the ninth century were evolving along parallel lines, but in those cases writers were combining prose with skaldic rather than eddic poetry in what would become the sub-genres of the kings’ sagas (konungasögur) and the sagas of Icelanders (Íslendingasögur).
A description of entertainment at a wedding feast at the Icelandic farm Reykjahólar in the year 1119 is found in the contemporary saga (samtíðarsaga) Þorgils saga ok Hafliða ‘The Saga of Þorgils and Hafliði’ (Brown 1952, 17-18, 72-3, 75-6), composed more than one hundred years after the event it purports to describe, and for that reason regarded with scepticism by some scholars (e.g. von See 1981c). If the colourful details of this description can be trusted (Clunies Ross 2010a, 18-20) – and one might think such detail would be unlikely to have been invented – at least one of the entertainments there was a fornaldarsaga about the legendary hero Hrómundr Gripsson, which included many stanzas (ok margar vísur með). There is no extant medieval version of this saga, but Hrómundr’s story survives in rímur (Griplur) from the fifteenth century and in a seventeenth-century fornaldarsaga version probably based on the rímur. The narrator of Þorgils saga adds that King Sverrir of Norway (r. 1177-1202) was entertained with this story, and it is further said that ‘he called such lying stories the most entertaining’ (ok kallaði hann slíkar lygisǫgur skemtiligastar). At the same time, the narrator states, some people are able to trace their genealogies to Hrómundr Gripsson, and presumably those people, like the descendants of the Hrafnistumenn, had good reason to consider the stories about their ancestors to hold at least a modicum of truth.
The description of the Reykjahólar wedding entertainment suggests that some form of the fornaldarsaga, incorporating both prose and verse, was being performed before an audience in the early twelfth century, presumably presented and transmitted orally. Another kind of evidence to support the idea that moves were afoot to link poetry and prose in a kind of legendary history comes from the presumed literary development during the twelfth and thirteenth centuries of the corpus of poetry now referred to as the Poetic (or Elder) Edda (cf. Larrington, Quinn and Schorn 2016). This term refers in the first instance to poems in alliterative verse with pre-Christian mythological and legendary subjects recorded in compilations with little or only a small amount of prose contextualisation. There are in fact only two such compilations extant, the more important being the late thirteenth-century Codex Regius of the Poetic Edda, GKS 2365 4°, of c. 1270, the other being the early fourteenth-century fragment AM 748 I a 4°, of c. 1325. Other poems of similar type were recorded in manuscripts containing versions of Snorri Sturluson’s Edda and in the historical compilations Hauksbók and Flateyjarbók (for details of the preservation contexts of poetry in eddic metres, see Clunies Ross 2016).
It has been argued (Lindblad 1954; 1980) that the Codex Regius collection of eddic poetry had a complex prehistory and may have been in process for some decades during the thirteenth century before it achieved the form in which we know it. The idea that collections of mythological and heroic poetry may have been brought together over a period of some years has also been extended to include a surmise that Snorri Sturluson may have had access to such a collection of mythological poems when he composed the Gylfaginning ‘The Delusion of Gylfi’ (Gylf) section of his Edda (SnE) in the 1220s. Drawing attention to the way in which fornaldarsögur such as Vǫls create a prosimetrum comprising a great deal of eddic poetry linked by means of a relatively superficial prose narrative, Anne Holtsmark (1965c) suggested that both the evolution of the fornaldarsaga and the gathering together of eddic poems in a compilation may have been parallel developments, with the origin of each lying in the desire to amplify in prose the unspoken but generally understood context of the action of these traditional poems, especially those based on heroic legend.
In the Codex Regius of the Poetic Edda, the explanatory prose links between individual poems and groups of stanzas (termed in German Begleitprosa) are much more numerous in the heroic section than in the first part of the collection devoted to mythological poems, where longer explanatory prose passages are confined to Grímnismál (Grí), Lokasenna (Lok) and Vǫlundarkviða (Vǫl). On the evidence of the somewhat sketchy provision of prose narratives in some of the fornaldarsögur that include heroic poetry, notably Vǫls, Hálf and Heiðr, it seems plausible to suggest that these sagas developed in much the same way as the Codex Regius collection, but that the impulse to strengthen narrative links between groups of stanzas and whole poems eventually brought about a more balanced kind of prosimetrum in which poetry and prose were better integrated.
Most of the extant manuscripts in which fornaldarsögur have been recorded are of quite late date, the majority of them being post-medieval. This makes it difficult to provide authoritative statements about the early history of the sub-genre and the ways in which it developed during the thirteenth and fourteenth centuries. It is also difficult to gauge the relative age of individual sagas, as their texts often exist in several different versions. However, it is clear, where we possess manuscripts of a particular saga of varying age, that both the poetry and the prose underwent change as time went on in response to the changing tastes and interests of the patrons for whom the sagas were being copied or reworked and the capacities of the scribes who wrote them down. It is possible to detect certain themes and changes of emphasis in later versions of a text like Ǫrv, for example, which point to new kinds of reception as the tastes of audiences changed (cf. Arnold 2010; Lassen 2009). Recent research into the fornaldarsögur has begun to investigate the nature of the sponsorship and likely audiences of these sagas and how the tastes and social expectations of their sponsors may have influenced their reworking into significant compilations (cf. Orning 2012; 2017). What is certain is that most fornaldarsögur retained and even increased their popularity in Iceland after the Middle Ages, and this is attested by the very large number of extant paper manuscripts from the fifteenth to the nineteenth centuries (for details see the web site Stories for All Time).
As far as the poetry in these sagas is concerned, it is also of varying age in many cases, and is also likely to have been reworked in the process of scribal copying. It is clear from the presence of late, non-metrical linguistic forms in many manuscripts that late- and post-medieval scribes often did not understand the requirements of Old Norse metres and introduced contemporary linguistic features into the texts they were copying or composing (see Sections 6 and 8 below). In several sagas certain groups of stanzas stand out clearly as individual long poems, as in Heiðr and Hálf, though not all the stanzas within these long poems are necessarily of equal age. In some cases metrical analysis (see Section 6 below) can reveal those parts of the verse component of individual sagas that are earlier than other parts. In other cases it can be seen that certain manuscripts, usually later ones, amplified their poetic texts with the addition of many more stanzas than earlier manuscripts contained. The most striking case of this kind is found in Ǫrv. By contrast, there are cases where one can see that a redactor is likely to have cut out stanzas that were in his exemplar. This is the case with the mid-seventeenth-century manuscript of Gautr, Holm papp 11 8°ˣ (papp11ˣ), which omits all but six stanzas of the poetry ascribed to Starkaðr, but retains in the prose text the introductory formula svá segir Starkaðr ‘so says Starkaðr’ in places where the stanzas would be expected to appear in the text, thus indicating that they were probably there in the copyist’s exemplar.