This is a backup server for Any changes made here will be lost.

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

4. The metres of skaldic poetry

4.1. Constitutive features of Old Norse and Germanic alliterative poetry (KEG)
4.2. Eddic metres (KEG)
4.3. Skaldic metres (KEG)
4.4. The metrical development of skaldic metres (KEG)
4.5. Formal structures in Old Norse poetry: stanzas and poems (KEG)

(Vol. I. Poetry for Scandinavian Rulers 1: From Mythological Times to c. 1035 > 2. General Introduction > 4. The metres of skaldic poetry > 4.1. Constitutive features of Old Norse and Germanic alliterative poetry)

4.1. Constitutive features of Old Norse and Germanic alliterative poetry (KEG)

The Old Norse metres are all rooted in the tradition of Germanic alliterative poetry, which in West Germanic territory is best exemplified by such poems as the Old English Beowulf, the Old Saxon Heliand and the Old High German Hildebrandslied. The fundamental structuring principle governing this poetry was alliteration, that is, the repetition of vocalic syllabic onsets and of phonetically similar consonantal onsets in two or three stressed positions in a long-line. Alliteration prescribes that all vowels can alliterate with each other (in Old Norse including [j]), but all consonants must be identical, that is, f- can only alliterate with f-, k- with k- etc., and the consonant clusters sk-, sp- and st- are treated as single phonetic units (see Gade 1995a, 4-5, 248 n. 15 and the literature cited there). A long-line falls into two half-lines divided by a metrical caesura. The first half-line (in Old Norse metrical terminology the ‘odd line’) can have one or two alliterative staves, and the second half-line (the ‘even line’) has one alliterative stave alliterating with the stave(s) in the odd line. Each half-line has four metrical positions, two fully stressed and two unstressed positions (or lesser stressed positions), and in the even lines the alliterative stave (hǫfuðstafr ‘main stave’) falls in the first stressed position. Consider the following half-stanza in fornyrðislag (Ív Sig 2/1-4II; the alliterative staves are emphasised):

Vas með jarli        afkárlyndum

vargs verðgjafi        vestr í eyjum

Was with the jarl        obstinate-minded

the wolf’s food-giver        west in the isles.

Alliteration falling on two distinct half-lines is found in runic inscriptions as early as the Gallehus horn (Run DR12VI, c. 450: Ek Hlewagastiʀ Holtijaʀ | horna tawido ‘I, Hlewagastir, the son of Holt (or ‘from Holt’), made the horn’), and must have been a part of the common Germanic heritage.

In Old Norse the root-syllable of a word was stressed more strongly than in the West Germanic languages, and therefore syncope (the loss of weakly stressed medial vowels) and apocope (the loss of weakly stressed final vowels) were carried through much more extensively in Old Norse than in West Germanic (compare, for example, OE sunu, ON sonr < Gmc *sunuz m. nom. sg. ‘son’; OE fremede, ON framði 1st pers. sg. pret. indic. ‘I did, furthered’). The same strong stress on root-syllables caused Old Norse unstressed prefixes to be reduced (OE gelīc, ON glíkr m. nom. sg. ‘like’) or lost altogether. The extensive syncope and apocope that took place c. 600-800 affected Old Norse poetry in that each metrical position is usually occupied by only one syllable. As in West Germanic poetry, a stressed position in fornyrðislag is typically filled by one long syllable or by two short ones (or a short plus a long) that are resolved, that is, in such lines as Gísl Magnkv 11/7II konungr ok jarlar ‘the king and the jarls’ and 17/3II austr við Elfi ‘east by the Götaälv’, konungr and austr are metrically equivalent and each occupies one metrical (stressed) position (for discussions of long and short syllables in Old Norse poetry, see Kuhn 1983, 53-5; Gade 1995a, 29-34). An unstressed position can contain one long or one short syllable, or two short ones, if they neutralise: in Ív Sig 15/7II þá vas svalt á sæ ‘then it was cold at sea’, the unstressed þá vas ‘then was’ metrically equals þvís‘which’ in 9/6II þvís vígði guð ‘which consecrated God’ (for resolution and neutralisation, see Kuhn 1983, 55-6; Gade 1995a, 60-6). An unstressed position can also accommodate two syllables if the second is elided, as in Gísl Magnkv 13/5II Stukku af almi ‘Flew from the elm-bow’ in which the a- in af is elided in hiatus (-u a-; see Kuhn 1983, 56-7; Gade 1995a, 66-7). Unlike West Germanic poetry, however, resolution can only occur in the first metrically fully stressed position (lift) in fornyrðislag lines, and an unstressed metrical position (dip) is not filled by more than two syllables. Hence Old Norse fornyrðislag mostly adheres to the same metrical patterns as West Germanic alliterative poetry, but the fillers of this metre are much pared down compared to those of its West Germanic relatives.

Metrically, odd and even fornyrðislag lines can be laid out as follows, according to Sievers’s five-type system.[15] The notations used below are the following:  is a long, fully stressed syllable,  a long syllable with secondary stress;  is a short fully stressed syllable,  a short syllable with secondary stress;  is a dip, and ̋ designates alliteration (syllables carrying alliteration are always fully stressed):

Odd lines:
Type A:     (A1)     (A2)     (A3)
Gísl Magnkv 4/5II Bjósk at brenna ‘Prepared to burn’
Gísl Magnkv 12/1II Margan hǫfðu ‘Many had’
Gísl Magnkv 2/5II Fylgðu ræsi ‘Accompanied the ruler’
A2k:        
Gísl Magnkv 2/3II lofðungr liði ‘the lord his troop’
Gísl Magnkv 1/7II Báleygs viðu ‘Báleygr’s trees’
Type B:        
Ív Sig 11/7II áðr Saxa sjǫt ‘before the Saxons’ dwellings’
Type C:     (C1)     (C2)     (C3)
Gísl Magnkv 15/5II Braut dýrr dreki ‘Broke the precious dragon’
Gísl Magnkv 5/3II gekk hár logi ‘swept the high flame’
Type D:     (D1)        
Gísl Magnkv 2/7II linns láðgefendr ‘the snake’s land-dispensers’
Gísl Magnkv 13/7II hvítmýlingar ‘white-muzzled arrows’
Gísl Magnkv 5/1II Hyrr sveimaði ‘Fire surged’
Type D4:        
Ív Sig 40/5II Raufsk ræsis lið ‘Scattered the ruler’s troop’
Gísl Magnkv 17/5II Liðskelfir tók ‘The troop-terrifier took’
Type E:        
Gísl Magnkv 15/3II steinóðr á stag ‘raging against the stay’
Even lines:
Type A2:    
Gísl Magnkv 1/4II landi ræna ‘the land rob’
A2k:    
Gísl Magnkv 3/2II siklings flota ‘the sovereign’s fleet’
Type B:    
Ív Sig 8/6II ok synðum hrauð ‘and the sins expiated’
Type C:     (C1)     (C2)     (C3)
Gísl Magnkv 1/8II með blm hjǫrvi ‘with the dark sword’
Gísl Magnkv 15/6II und Dana skelfi ‘beneath the Danes’ terrifier’
Type D:    
Gísl Magnkv 15/8II hafs glymbrúði ‘the ocean’s roaring-bride’
Type E or D4:        
Gísl Magnkv 9/4II Ívistar gram ‘North Uist’s lord’

As this overview shows, fornyrðislag is predominantly characterised by four-syllabic half-lines. To be sure, poetry composed in fornyrðislag included in the Poetic Edda cannot by any means be called syllable-counting, but the tight, minimal metrical patterns clearly paved the way for and facilitated the development that resulted in the stylised, syllable-counting skaldic metres. The paragraphs below discuss the eddic and skaldic metres in more detail, including the poetry composed in the respective metres and metrical innovations that developed over time.

© 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.

This is a backup server for Any changes made here will be lost.