Íslenskt mál og almenn málfræði - 2020, Side 326

example of both strategies may be found in the aforementioned verse skrǫkberǫnd -
um (Arbj 2.4), with natural secondary stress on both the second and the third sylla-
ble, of which only the first carries a lift (-ber-), whereas the other is placed in a dip
(-ǫnd-). That in ordinary language the third syllable was not completely unstressed
may be seen from its full vowel ǫ, cf. the opposite in an inflectional form like kǫst -
uðu ‘(they) threw’ for older or expected *kǫstǫðu < *kastaðu, where the second ǫ
caused by u-mutation cannot be maintained in an unstressed syllable and thus
enters into the un stress ed vowel system as u.
Þorgeir’s assignment of metrically strong positions to naturally long syllables
comes into conflict with the alliterative patterns of Old Norse poetry as well.
Traditionally, alliteration has been seen as a device for emphasising positions
already stressed within a verse. From this it follows that only stressed positions
— and indeed the most prominent of those (cf. Rieger’s (1876) proposed hierar-
chy) — may carry alliteration. Since there are many examples of short-stemmed
nouns carrying alliteration in Old Norse poetry, Þorgeir is forced to deny the
connection between strong positions and alliteration. For instance, the verses á
hlið aðra and þótt fé eigi (Arbj 10.4 and 20.4), are scanned “vvsx” with a “weak” sec-
ond position (p. 132), even though this position in both instances carries the allit-
eration (here marked in bold). These verses are exceptional in having a short
stressed syllable before a word beginning with a vowel (fé counts as short in Old
Norse), and as Þorgeir points out, they seem to be peculiarly Egill’s, but one is
still puzzled by the treatment of the first lift as “weak”, not least in light of Þor -
geir’s later treatment of short-stemmed alliterating nouns in odd verses, such as
hǫfuðlausn or vinar míns (Arbj 8.7 and 15.5), which are analysed as having a
“strong” first position (in, e.g., the scansion on p. 159). This does not seem very
consistent. He rarely comments on the interplay between stress and alliteration,
but on pp. 132f., we find this statement: “Usually, alliteration is assumed to be
only on syllables in lifts, but I claim that the poetic evidence supports rather that
it is always in a heavy syllable, or syllables that can be made heavy by cohesion,
or on the first of two syllables in a Verschleifung.” Since almost any short syllable
can be made heavy by cohesion, by placing it before a word or word-element
beginning in a consonant — as in the verse cited by Þorgeir, á gjafstóli (Hkv 34.4;
p. 132) — this does not solve the problem.
Linguistics
Þorgeir Sigurðsson has generally a good grip on skaldic language and linguistics.
I have noted only some minor problems:
(1) In the reconstructed text of st. 15 (p. 191), Þorgeir deletes the opening
word Nú, despite acknowledging in the notes that this is not necessary if
Klaus Johan Myrvoll326