Íslenskt mál og almenn málfræði - 01.01.2011, Side 74

72 Ragnar Ingi Aðalsteinsson og Sigurður Konráðsson
with sl-, sn- with sn- and sm- with sm-), as sp-, st- and sk- had always done. (A few poets also
make use of the so-called epenthetic alliteration, where sl- and sn- alliterate with st-
(because of the epenthetic [t] that has snuck into the sl- and rw-clusters) and sm- with sp-
(because of the epenthetic [p] in the rtw-cluster).) 5-alliteration then reappears in the l8th
and I9th century and some poets used it until the loth century, but now it seems to have
been completely abandoned again. These changes in alliteration of the initial clusters sl-,
sm- and sn- are discussed in this article. The first part of the paper focuses on the devel-
opoment and changes in this alliteration in poetry. In the second part these changes are
compared to the relative frequency of the relevant initial clusters in eight prose texts from
the I3th to the 2ist century. The idea is to investigate if changes in the frequency of allit-
eration with these clusters corresponds to changes in their relative frequency in prose
texts. In alliteration, all these clusters virtually disappear around 1400 and are not seen
after that except as gnystuduls (i.e., alliterating with themselves) until the i-alliteration is
revived in the i8th century. In prose, on the other hand, the relative frequency of these
clusters appears to be similar throughout the whole time frame. No specific changes
appear. The conclusion seems to be that the changes in the use of these clusters in alliter-
ation around 1400 must be caused by changes in their pronunciation and/or rules having
to do with alliteration in poetry and not by changes in the vocabulary. — In addition, it
appears that before 1400 alliteration with sm- is quite limited. When í-alliteration of sl-,
sn- and im-clusters comes into use again, mostly during the i8th century, the number of
words with sm- as an initial cluster in alliterating words increases considerably. This can
be considered an indication that sm- was a gnystudul in ancient times, only alliterating with
itself and not with other initial clusters nor with i+vowel. Hence its occurrence was nec-
essarily limited in alliteration.
Ragnar Ingi Aðabteinsson
Menntavísindasviði Háskála Islands
Stakkahlíð
IS-105 Reykjavík, ÍSLAND
ria@hi.is
Sigurður Konráðsson
Menntavísindasviði Háskóla íslands
Stakkahlíð
IS-105 Reykjavík, ÍSLAND
sigkon@hi.is