Íslenskt mál og almenn málfræði - 01.01.2013, Side 219

On Haukur Þorgeirsson’s doctoral dissertation
217
fur Þórðarson (around 1210—1259). Haukur follows Myrvoll and Skomedal’s
view stating that the TGT is strongly influenced by Latin models, particularly
Priscian and Donatus, but becomes more independent in its course and starts to
reflect tonal distinctions. In particular, 2banum would metri causa stand for
intended ^b&num (p. 217). Thus, banum (dat. pl. from ban ‘prayer’ < early Olcel.
bdn) has the type of toneme 2 whereas btz-num (from bár ‘farm’, with the article
-num suffixed to it) receives toneme 1. Haukur argues that the Icelandic and
Norwegian language were so closely related in the 13* c. that the Norwegian
tonal distinctions are likely to be found in Icelandic as well (p. 217-218). As this
is probably the most solid and original contribution of the dissertation, I do not
refrain from citing the author at some length (p. 219, my emphases):
Ætla má að í íslenskum kveðskap síðmiðalda hafi verið rík hefð fyrir að taka
tillit til tónkvæðis þegar kom að orðmyndum eins og 'góður eða 2móður. Til
að orð gætu rímað saman þurftu þau að hafa sams konar tónkvæði [...] Til
viðbótar er hefðarregla sem segir að orð sem enda á -ur en hafa tiltekinn
yfirsneiðarlegan þátt [i.e. suprasegmental features; M.S.] eigi að meðhönd-
last eins og þau endi á -r í lok vísuorðs. I norsku og sænsku hafa orð eins og
2heder, sem til eru orðin úr einkvæðum orðum, svipað tónafar og eins-
atkvæðisorð. Við getum hugsað okkur að sama hafi gilt í íslensku og að það
hafi gert fyrrnefnda hefðarreglu eðlilega.
In the present opponent’s view, this is the strongest and most interesting part of
the doctoral thesis reflecting the author’s analytical procedure and his sound
methodology. Haukur uses two stanzas in the ferskeytt metre to exemplify the
distribution of old r-endings versus «r-endings (with the epenthetic vowel insert-
ed). The definite article is analyzed in similar terms and it is demonstrated that
rhyme of the type 2alinn (past participle) — ^dal-inn (acc. of dalr, with the article
-inn suffixed to it) is particularly rare (pp. 218—224). Haukur formulates a
requirement or type trait T (= differentia specifica) which states that the rhyming
words are originally disyllabic and that they display toneme 2 (p. 224):
Hamla T: Kvenrím í ferskeyttum hætti er með þeim hætti að tvö tvíkvæð
orð ríma saman og verða bæði að hafa seinna tónkvæði.
The author concludes that tonal distinctions disappeared in Icelandic around
1600 when skalds no longer made a distinction between rhyme pairs such as
2fengið — 2mengið which fulfill the requirement T, and 2búna (past participle, acc.
sg. fem. from búinn) — Hrú-na (with the suffixed article) which do not (p. 226).
The author is aware of stylistic issues that influence the development. Not-
withstanding problems with the use of epenthetic vowels in the particular hálf-
hnepptr metre, the results in chapter 9.12 may be regarded as the cornerstone of
the whole dissertation.