Íslenskt mál og almenn málfræði - 01.01.2021, Qupperneq 23

Line 5.3 appears not to be a clause-line, which is due to an editorial mis-
take, corrected by Myrvoll (2014:351) with a reference to Íslenzk fornrit
VI, 48. The line has rhyme in all its manuscripts. Thus, 21 lines are with-
out rhyme or 16% of the odd lines in Gísli’s lausavísur. Ten of these lines
are the third line of a half-stanza. In Gísli’s lausavísur, around 60% of the
third lines are clause-lines. The likelihood of picking ten clause-lines by
coincidence from these is less than 1%.29

4. Haustlǫng versus proto-dróttkvætt
Now that I have, in Section 3, established that lack of rhyme in odd lines
does not distinguish Haustlǫng from younger poems in the 10th century,
the question arises whether the poem has other poetic features that do. I
will show in the following that it does not.
I divide dróttkvætt poetry into three groups. The first group, I call
proto-dróttkvætt. The only poem in this group is Ragnarsdrápa by Bragi
Boddason. It seems that features of proto-dróttkvætt lasted longer in
lausa vísur than in poems. Five eight-line lausavísur by the Orcadian Earl
Torf-Einarr from around the year 90030 have features of proto-dróttkvætt,
and some lausavísur by Egill Skallagrímsson and by Kormákr Ǫg mundar -
son as well, while Gísli’s lausavísur almost lack them completely.31 The
second group, I call early dróttkvætt. Most poetry from the 10th century
belongs to this group, and I argue that so does the poem Haustlǫng. The
third group, I call classical dróttkvætt. In this group is the dróttkvætt
poetry that Snorri Sturluson described and prescribed for his readers. I
use the year 1000 as a dividing line between classical and early dróttkvætt.
Snorri Sturluson gives a description of obsolete features of dróttkvætt
in stanzas 54–58 of Háttatal (Faulkes 2007:24–26). He explains the
abnormality of each stanza and attributes them to old poets, including
Torf-Einarr, Bragi, and Egill. These features are the following:
1 Odd lines in proto-dróttkvætt did not need any rhyme. I have al -
ready discussed this item.
2 Even lines did not need full-rhyme (they normally had half-rhyme
Haustlǫng 23
29 If the likelihood for picking one clause-line by coincidence is 0.6, the likelihood for
doing so ten times is 0.610 = 0.006 which is 0.6%.
30 I use Finnur Jónson’s dating in his 1912–1915 dróttkvætt edition.
31 An exception is in Gísli Súrsson’s lausavísa 28.6 í hjǫrregni þvegnar, with alliteration
on hjǫr- in the second metrical position of an even line (see item 3).