Íslenskt mál og almenn málfræði - 01.01.1996, Side 194

192 Kristján Arnason
Riad, Tomas. 1992. Structures in Germanic Prosody. A Diachronic Study with Speci-
al Reference to the Nordic Languages. Stockholm University, Department of
Scandinavian Languages, Stokkhólmi.
Trommelen, Mieke & Wint Zonneveld. 1994. Word Stress in English, Dutch, Germ-
an, Norwegian and Swedish. Handrit, Háskólanum í Utrecht, Hollandi.
Þorsteinn G. Indriðason. 1994. Regluvirkni í orðasafni og utan þess. Málvísindastofn-
un Háskóla Islands, Reykjavík.
SUMMARY
This paper explores and compares the stress patterns of Icelandic and Faroese. The
main object is to investigate how the indigenous left strong pattern and the right strong
pattern of many foreign words, particularly in Faroese, can coexist.
Section 2 gives an overview of the stress patterns of Icelandic, and section 3 descri-
bes the situation in Modern Faroese. It is shown that although the main rule is for
Icelandic words to have initial stress, there are interesting exceptions, where it must be
assumed that forms that from the morphosyntactic point of view must be seen as words,
nevertheless follow the right strong pattern of postlexical constructions. A similar phe-
nomenon is observed in native Faroese words, and furthermore, the vast majority of
loanwords have non-initial stress. (Most loans in lcelandic follow the native pattern.)
The right strong pattern is thus much more common in Faroese than in Icelandic.
The account proposed in this paper is that when loanwords are stressed in the
European way in Faroese (and to the extent that this happens in Icelandic as well), the
forms are treated as pseudo-compounds, which means that they are analysed into feet
as if they were native compounds, and then they are stressed according to the right
strong pattem, which is quite commonly used in native compounds anyway. Thus
loans like Faroese hexa'meter ‘hexameter’, peli'kanur ‘pelican’ and disko'tek ‘disco-
teque’ have the same pattern as native compounds like buröar'vektir ‘birth weights’
and harum'framt ‘furthermore’. To the extent that foreign forms like Co'rolla and
Ge'valia occur in Icelandic, it is suggested that they be treated in the same way.
The unanswered question is by which criteria the pseudostructure is assigned, in
other words what becomes a stress foot and what not. Obviously, this must to a large
extent be determined by the stress pattern which is being imitated, but in certain cases
the stress pattern of a borrowing neither follows the native pattern nor that of the lend-
ing language, which suggests that there may be some special principles at work. Tliis
problem is left for further research.
Kristján Arnason
Háskóla Islands
Arnagarði v. Suðurgötu
IS-101 Reykjavík, ÍSLAND
kristarn@rhi. hi. is