Edith Marold 2017, ‘Snorra Edda (Prologue, Gylfaginning, Skáldskaparmál)’ in Kari Ellen Gade and Edith Marold (eds), Poetry from Treatises on Poetics. Skaldic Poetry of the Scandinavian Middle Ages 3. Turnhout: Brepols [check printed volume for citation].
Snorra Edda (SnE) is the name given to a collection of several texts which have different authors and originated at different times. All complete mss (R, Tˣ, W, U; but not the fragments A, B, C) contain a Prologue followed by Gylfaginning ‘the Delusion of Gylfi’ (Gylf), Skáldskaparmál ‘the Diction of Poetry’ (Skm) and Háttatal ‘Enumeration of Verse-forms’ (Ht). Ms. W has an additional four grammatical treatises with a unique Preface to them, which are inserted between Skm and Ht. Snorri’s authorship of SnE is confirmed by a remark at the beginning of ms. U (SnE 1848-87, II, 250): Bok þessi heitir edda. hana hevir saman setta snorri sturlo sonr ‘This book is called Edda. Snorri Sturluson has compiled it’. In the fragmentary ms. A (SnE 1848-87, II, 427-8), Snorri is mentioned as the author of Skm. Snorri’s authorship is generally accepted for Gylf, Skm and Ht, but some scholars dispute his authorship of the Prologue (e.g. Heusler 1908, 12; von See 1988, 29; 1990, 122; Clunies Ross 1987, 12, on the other hand, regards the prologue as an important key to the conception of the Edda – for further literature see there). The first, second and fourth of the grammatical treatises are anonymous; the third is the work of Snorri’s nephew, Óláfr Þórðarson. The First Grammatical Treatise (FGT) contains one helmingr and a couplet, both edited in other volumes of SkP (Ótt Hfl 8/5-8I, ÞjóðA Har 3/1-2II) while The Second Grammatical Treatise (SGT) has no stanzas; hence only the poetry from The Third Grammatical Treatise (TGT) and The Fourth Grammatical Treatise (FoGT) is edited in the present volume (see Sections 4.2.5 and 4.2.6 below).
A. The Prologue
Even though Snorri’s authorship of the Prologue cannot be established with any certainty, it nevertheless serves as a preface to Gylf. The first part of the Prologue deals with the origin of religions and the multitude of languages, and the second part relates the immigration of the Æsir, the inhabitants of Troy descended from King Priam, to Scandinavia. Led by Óðinn, they migrate to the north (Scandinavia and Saxland, northern Germany) and establish kingdoms there. This section of the Prologue corresponds roughly to the beginning of Heimskringla (Hkr), but the two versions are not fully identical. The story of the Swedish king Gylfi, who is cheated out of land by Gefjon, one of the Æsir, serves as a bridge between the Prologue and Gylf (see Bragi Frag 1 and Introduction there). King Gylfi is defeated several times by the Æsir when competing with them in sorcery, and he decides to explore the background of their magic powers.
B. Gylfaginning
Gylf is a presentation of the pagan Scandinavian religion in a dialogue between Gylfi, who travels under the alias Gangleri, and three of the Æsir who appear to him in a sort of visionary illusion (sjónhverfing) as Hár (the High One), Jafnhár (the Equally High One) and Þriði (the Third) in their stronghold. In Gylf Snorri uses two forms of dialogue. The first is the knowledge contest which is well-known from eddic poetry, e.g. Vafþrúðnismál (Vafþr). Snorri does not follow this model fully, however, because the second part of such dialogues, when there is a reversal of roles and the interrogator becomes the one who is questioned, is lacking. In Gylf, Gylfi remains the interrogator until the very end. The model of contest is abandoned very early in the first part of Gylf, and from then on the conversation resembles more the dialogues between teacher and student which are known from the Latin tradition and were taken over into the Old Norse literary tradition, e.g. in Konungs skuggsjá and Elucidarius. Gylf also deviates from that pattern, however, because the intention of the Æsir to persuade Gylfi to accept the religion presented to him finds clear expression in the dialogue.
The conversation proceeds from the presentation of the Æsir, beginning with Alfǫðr, who is later identified as Óðinn, to the creation of the world from the body of the slain giant Ymir, the construction of Ásgarðr, the stronghold of the Æsir, and the fettering of the wolf Fenrir, which is described in great detail. Here, Gylfi professes his faith to Óðinn/Alfǫðr. Some of the adventures of Þórr follow – the fishing for Miðgarðsormr (‘the World Serpent’), the journey to Útgarðaloki etc., and finally the description of ragnarǫk, the end of the world. After that the Æsir finish the conversation by dispelling the illusion. Gylfi finds himself in an open field, and later on he begins to spread his new faith. The Æsir decide to identify themselves with the Æsir of the mythical stories they had told Gylfi.
There are likely to have been two reasons, not mutually exclusive, why Snorri did not write a systematic presentation of the pagan religion, but rather presented it in the form of a dialogue. First of all, he used the existing model of dialogic teaching that characterised the learned Latin and Icelandic tradition, which he combined with the vernacular model of gnomic poetry known from the Poetic Edda. Secondly, the pointed presentation of Gylf as ‘a delusion of Gylfi’ which identifies the myths as belonging to the fabulae, narrations one should not believe in, but which poets created for the entertainment of their audience, gave Snorri the opportunity to distance himself from the pagan religion.
The interpretation of Gylf has been extremely controversial and ranges between the polar opposites of a confession of faith and a critique of pagan religion. Kuhn (1942, 163-4; 1967, 755), for example, interprets Hár’s words Ok þat er mín trúa ‘And that is my belief’ (SnE 2005, 11) as an Ausbruch religiöser Verehrung ‘an outburst of religious adoration’ by Snorri Sturluson, while other scholars see him as presenting the pagan religion as the result of deception by devilish or demonic forces (Baetke 1950; Holtsmark 1964; Weber 1986a; 1986b; for a moderate view of pagan religion as a historical and cultural phenomenon see e.g. Dronke and Dronke 1977; Clunies Ross 1987; 1992b; von See 1988; 1990; 1993; Faulkes 1983; Beck 1993; 1994b; Marold 1998c; SnE 2005, xxvii-xxix). What is clear, however, is that Gylf presents the myths that are necessary for the understanding and use of mythological kennings. It is remarkable that, with two exceptions (Bragi Frag 1, Þhorn Harkv 11I), no skaldic stanza, but only eddic ones are cited in Gylf. Hence Gylf is not a compendium of skaldic poetry, but it serves as the background for many mythological ideas which are found in skaldic poetry.
The third part of SnE, Skm, presents the most striking linguistic features of skaldic diction, namely, kennings and heiti as variation of or substitution for plain language (see the General Introduction in SkP I, lxix-lxxxix). The two categories of kennings and heiti determine the structure of Skm. This is most clearly shown by ms. U, in which a differentiation between kent ‘paraphrased’ and ókent ‘not paraphrased’ is found already in the introduction, whereas the other mss (R, Tˣ, W) introduce three kinds of poetic expressions, kenning, heiti and fornǫfn (the last category is difficult to define; see the General Introduction in SkP I, lxxiv-lxxv). The source of these three kinds of poetic expressions, which are also found in the prose commentary on Ht, could possibly be medieval Latin poetics (see the General Introduction in SkP I, lxxv). However, the three categories do not have any effect on the bi-partite structure of Skm, which clearly reflects the duality of kenning and heiti, although some kennings are found in the section on heiti (see below).
In Skm Snorri maintains the presentation of his material in dialogue form as in Gylf. Longer prose accounts of myths like that of Hrungnir or Geirrøðr are not part of the dialogue, but even there Bragi is the narrator (on narratives in SnE see Clunies Ross 1998d). The beginning of Skm, the so-called Bragarœður ‘the Speeches of Bragi’, establishes a scene of conversation at a mythological level; like Gylfi, a man called Ægir comes to the Æsir and is received with visionary illusions (sjónhverfingar) like his predecessor in Gylf. At the feast in the evening Ægir is sitting beside Bragi and they begin to converse about poetry. Ægir is the interrogator, Bragi the expert. This conversation is not a contest as in Gylf, but follows a simple question and answer model. It is curious that Ægir, who appears as a sea-giant in other places, takes on a role similar to Gylfi’s.
Before the conversation turns to the language of poetry, the myth of the mead of poetry is told, which provides the mythical background and foundation of poetry. However, it has been debated (Frank 1981) whether this myth might not be a (re)construction by Snorri based on his interpretation of the numerous kennings for ‘poetry’ which use this myth as a foundation.
The myth itself is the story of a precious drink which changes owners repeatedly – dwarfs, giants, Óðinn. It was created from the blood of Kvasir, a mythical being born of the mixed saliva of the Æsir and Vanir, which they spat into a vessel as they made peace at the end of a prolonged war. The dwarfs killed Kvasir and created the mead of poetry from his blood mixed with honey, but they lost possession of it when they had to hand it over as a ransom to the giant Suttungr whose daughter-in-law they had killed. Suttungr extorted the mead from the dwarfs by exposing them on a skerry where they would have drowned during high tide. Suttungr kept the mead in three vessels inside a mountain, guarded by his daughter Gunnlǫð. Óðinn gained access to her, and they agreed that he should be granted one draught from each of the three vessels after sleeping three nights with the giantess. However, he emptied the vessels completely and escaped, flying back to the stronghold of the Æsir in the shape of an eagle and spitting out the mead into containers.
After this story has been told, the actual theme of Skm, the description and explanation of poetic language, begins. The conversation between Ægir and Bragi opens very systematically by outlining the two main principles of poetic composition, mál ‘language’ and hættir ‘verse-forms’, followed by the special theme of Skm, the designation of things by means of kennings and heiti.
Yet before the presentation of the kennings starts, an authorial insert by Snorri interrupts the conversation between Bragi and Ægir, reverting back to the theme of Gylf (and possibly also to the Prologue). This is Snorri’s well-known dedication of his work to ‘young poets’, which at the same time is a warning against the pagan religion that arose from the alleged falsification of the history of Troy by the Æsir (SnE 1998, I, 5). Nevertheless the old kennings should not be forgotten or dismissed, according to Snorri. This section is called Eptirmáli ‘Epilogue’ in SnE 1848-87, owing to the editors’ assumption that both the preceding so-called Bragarœður and Eptirmáli belonged to Gylf. Later editors regard it as an insertion in Skm, and its presence could have been prompted by the fact that it is followed directly by a long section in which kennings for all the gods are dealt with, together with several myths explaining these kennings.
The conversation between Ægir and Bragi now changes to the typical teacher-student format, i.e. a series of questions and answers. Stereotyped questions such as hvernig skal kenna X ‘how should one paraphrase X’ are answered by enumerations of kennings for each referent, in most cases illustrated by examples from skaldic poetry. These are often only helmingar or single stanzas rather than whole poems, and the names of poets are given (though some of the poetry is anonymous) but the titles of the poems from which the stanzas are taken are usually not indicated. Larger parts of poems or whole poems are rarely cited, and when they are, it is mostly in connection with the narration of myths or heroic legends. Such poems or parts of poems are cited to explain the origin of kennings or when they are necessary for the understanding of the underlying myths (see also Section 2 above).
Some of the first myths told in Skm are of Þórr’s fights against the giants Hrungnir and Geirrøðr, which end with the citation of the corresponding parts of Þjóðólfr ór Hvini’s Haustlǫng (Þjóð Haustl 14-20) and the entire Þórsdrápa by Eilífr Goðrúnarson (Eil Þdr). The presentation of kennings for the goddess Iðunn is followed by the first part of Haustl (sts 1-13). The very extensive section on gold-kennings contains a great deal of mythological material and many heroic legends. The gold-kenning ‘fire of Ægir’, for example, is said to derive from the description of the source of illumination in the hall of Ægir, the sea-giant, and from that the pattern ‘fire of the sea’ for ‘gold’ is generated. The gold-kenning ‘Sif’s hair’ is explained by the story of Loki cutting off Sif’s hair and combined with the origin of the treasures of the gods. The kenning ‘otter-payment’ or ‘compensation for the otter’ prompts the retelling of the complete legend of the Nibelungs (ON Niflungar), beginning with the mythical origin of their treasure and culminating in the deaths of Guðrún’s sons Hamðir and Sǫrli in the hall of Jǫrmunrekr. The prose narration of this legend is followed by sts 3-7 of Bragi’s Ragnarsdrápa (Bragi Rdr). The gold-kenning ‘Fróði’s flour’ is explained by the story of the giantesses Fenja and Menja grinding gold for King Fróði of Denmark, after which the entire Grottasǫngr (Grott) is cited. The narration of Hrólfr kraki’s battle against the Swedish king Aðils and their subsequent dealings at Fýrisvellir, when Hrólfr ‘sowed’ gold on the ground to delay the pursuit of the Swedes, serves as an explanation of the gold-kennings ‘seeds of Kraki’ and ‘seeds of Fýrisvellir’. The kenning ‘storm of the Hjaðningar’ leads to the narration of the legend of Hildr and the eternal battle of the Hjaðningar, which is illustrated by the corresponding section of Rdr (Bragi Rdr 8-12). The second section of Skm, which is dedicated to the heiti, does not contain any prose narratives.
The order of the subject matter within the two parts of Skm on kennings and heiti follows a particular structure which is essentially the same for kennings and heiti, even though it differs in detail (see below). The structure of both groups, kennings and heiti, consists of sections on the concepts ‘poetry’, ‘gods’, ‘cosmos’ and ‘humans’. However, within these groups there are considerable differences.
(a) The section on poetry gives numerous examples of kennings based on the myth of the mead of poetry outlined above, whereas the heiti section only gives examples for the designations for ‘poetry’ which are listed at the beginning of the section.
(b) The section on the kennings for gods begins with Óðinn; after that kennings for the other male gods are given, followed by kennings for the goddesses. Óðinn and Þórr receive a detailed treatment, also with narratives (in particular, stories about Þórr’s exploits); the other twelve male gods are allotted less space. Sample stanzas containing kennings are only given for Njǫrðr, Freyr and Loki. Of the four goddesses Frigg, Freyja, Sif and Iðunn, only Iðunn receives a lengthier treatment, most likely because there was a section of Haustl (sts 1-13) devoted to her abduction which contained kennings for her and could be cited as illustrations of these.
The section of the heiti for gods is relatively short and only the general terms for the pagan gods, such as bǫnd, hǫpt, rǫgn etc. are given; each term is followed by one stanza that illustrates the use of the word in poetry. No heiti are given for the individual gods.
(c) The cosmological sections on kennings and heiti treat – in differing order – sky, sun, the heavenly bodies, earth, sea, air and wind, fire, the times of the day and the seasons. A significant difference between the kenning-section and the heiti section is that terms for animals like wolf, bear, stag, horse, oxen, snake, cattle, sheep, swine, raven and eagle have been included in the heiti section. In some cases these terms comprise not only heiti but also kennings, however, and it is possible that this group had been accidentally omitted when the kenning-section was written.
(d) The sections on kennings and heiti for humans are also very dissimilar. Both give terms for ‘man’ and ‘woman’ and for several social groups, though in different order. The heiti-section contains the much-discussed paragraph about viðkenningar, sannkenningar and fornǫfn (see the General Introduction in SkP I, lxxiv-lxxv). Whereas the heiti-section only gives terms for body parts and mental qualities, the kenning-section contains a large number of concepts which were used as determinants in kennings for ‘man’, such as ‘gold’, ‘battle’, ‘weapon’, ‘armour’ and ‘ship’.
A look at the content and structure of Skm makes it obvious that SnE was a work in progress. Faulkes (SnE 1998, I, xi) thinks that the work was unfinished when Snorri died in 1241. Yet it is also possible that interpolations were made later by others. There are not only differences between the individual mss of SnE, but the structural plan is not consistently carried out. In the heiti-section in the second part of Skm, kennings and heiti are sometimes mixed, and one gets the impression that items omitted in the kenning-section have been added here.
The sources for Skm are manifold. Most of the skaldic poetry cited in SnE probably derives from oral sources, but it is possible that Snorri had lists of kennings and heiti which had been collected already in the twelfth century (see Section 4.2.3 below).
There are considerable differences between the redactions of Skm in mss R, Tˣ, W, U, A, B and C. Ms. U, in particular, differs from the other mss because the category fornǫfn is lacking and it only distinguishes between kent ‘paraphrased’ and ókent ‘not paraphrased’. All references to the Trojan War are missing in that ms. as well, and some longer quotations from Haustl and Rdr have been omitted, which is also the case with Þdr (for a complete description, see SnE 1998, I, xxxix-xliii). Scholarly opinion on the status of the U version in the compilation of SnE differs. Some scholars regard it as Snorri’s initial version, while others view it as an abbreviated version of the longer redactions found in mss R, Tˣ and W (see SnE 1998, I, xli-xliv; U 2012; for an overview of earlier evaluations of U, see Zetterholm 1949; Sävborg 2013).
The mss differ in terms of their value as sources for our modern knowledge of skaldic poetry, and the number of skaldic stanzas transmitted in the seven mss of Skm also varies: ms. R has 470 stanzas, Tˣ 463 stanzas, W 255 stanzas, U 249 stanzas, A 355 stanzas, B 308 stanzas and C has 278 stanzas.
There are other factors which affect the value of SnE as a source for the study of skaldic poetry. Stanzas or helmingar are seldom assigned to named poems, though it is likely that they originated as parts of longer poems. Hence it is incumbent on an editor to decide whether they belong to a certain poem; e.g. of the ten dróttkvætt stanzas attributed to Þjóðólfr Arnórsson’s Sexstefja which are preserved in SnE only, seven can safely be assigned to that poem (see Introduction to ÞjóðA SexII). Sometimes single stanzas have been assigned to specific poems in earlier editions, yet later research has cast doubt on these reconstructions (see e.g. Introduction to KormǪ Sigdr). Faulkes (SnE 1998, I, xv) rightly observes that ‘many of the attributions in Skj are based on guesswork’. Another problem is that sometimes neither the nickname nor the patronymic of a poet is given. Thus ‘Einarr’ could be either Einarr skálaglamm or Einarr Skúlason, and ‘Þjóðólfr’ could be either Þjóðólfr ór Hvíni or Þjóðólfr Arnórsson (see the discussion of this problem in Section 9 below).