Íslenskt mál og almenn málfræði - 2014, Síða 119

bóka félagið, Reykjavík. [Fáanleg sem vatnsmerkt pdf-skjal í vefverslun Forlagsins, sjá
http://www.forlagid.is/?p=623043.]
Kristján Árnason. 2011. The Phonology of Icelandic and Faroese. Oxford University Press,
Oxford.
Kristján Árnason. 2013. Stíll og bragur. Um form og formgerðir íslenskra texta. Hið íslenska
bókmenntafélag, Reykjavík.
Magnús Pétursson. 1976. Drög að almennri og íslenskri hljóðfræði. Iðunn, Reykjavík.
Magnús Pétursson. 1978. Drög að hljóðkerfisfræði. Iðunn, Reykjavík.
Ragnar Ingi Aðalsteinsson. 2010. Tólf alda tryggð. Athugun á þróun stuðlasetningar frá elsta
þekktum norrænum kveðskap fram til nútímans. Doktorsritgerð. Hugvísindastofnun
Háskóla Íslands, Reykjavík.
Schulte, Michael. 2013. On Haukur Þorgeirsson’s Doctoral Dissertation. Íslenskt mál 35:
213–223.
Stefán Karlsson. 1989. Tungan. Frosti F. Jóhannsson (ritstj.): Íslensk þjóðmenning VI.
Munnmenntir og bókmenning, bls. 1–54. Þjóðsaga, Reykjavík.
summary
‘Alliteration, poetic traditions and phonology’
Keywords: alliteration, phonemes, phonological rules, equivalence classes, poetic tradi-
tions
This paper discusses some issues concerning the relationship between alliteration, phono-
logy and phonetics. It is a continuation of a discussion in the last issue of Íslenskt mál in a
report on Haukur Thorgeirsson’s doctoral defense.
Most of the discussion centers around the alliteration of palatal and velar consonants.
It had previously been argued (by Höskuldur Thráinsson 1981) that palatal stops alliterate
with velar stops because all palatal stops are derived from underlying velar ones. Haukur
argues in his dissertation that this cannot be true because there are too many exceptions to
proposed palatalization rules for it to be possible that they are genuine phonological rules.
It is argued in this paper that the exceptions do not matter — it is still possible to argue
that all palatal stops are derived from velar ones and this is all that matters with respect to
alliteration.
The paper also briefly reviews arguments deriving hj-, hl-, hn- and hr-words from
representations containing an initial /h/ since words of this kind alliterate with each other
and with words beginning with an /h/ followed by a vowel. But since [h] does not always
show up in the hj-, hl-, hn- and hr-words (possibly never in hj- words), according to
acoustic measurements, it has been maintained that this kind of analysis is too abstract.
Besides, it would raise the question what would prevent a similar analysis of words be -
ginning with voiceless fricatives like [f] and [θ] (i.e. taking them to represent underlying
/hv/ and /hð/). This paper points out, however, that the alliteration evidence argues
against such an analysis (f-words and þ-words to not alliterate with h-words). In addition,
voiceless fricatives like [f] and [θ] (and voiceless obstruents in general) are extremely com-
mon and natural whereas voiceless sonorants are a very rare and marked phenomenon.
Stuðlar, hefðarreglur, hljóðkerfi 119