Íslenskt mál og almenn málfræði - 01.01.1985, Blaðsíða 131

Narrative Inversion in Old Icelandic
129
and so on in her calculations. However, Sigurðsson argues, narrative
inversion is a declarative clause phenomenon, and it is the order of the
subject and the finite verb in declarative clauses which should consti-
tute the basis for the classification of languages according to word-
order types. Since direct yes/no-questions usually have the finite verb
in first position in Old Icelandic, we cannot rely on Kossuth’s fre-
quency figures when we want to know how frequent the narrative in-
version construction is in the Sagas. Kossuth’s figures must be too
high, Sigurðsson concludes. The result he gets from his own frequency
study of word order pattems confirms this suspicion: in the seven OI
texts he investigates, the word order VS is found in 11% of the declara-
tive clauses, to be compared with the 16% reported by Kossuth. Al-
though the difference does not seem to be dramatic, Sigurðsson takes it
as one of his reasons for rejecting Kossuth’s SV- KS-hypothesis. Notic-
ing that the VS-order is rare in subordinate clauses, that it is not con-
sidered as a neutral word order, and that it is bound to certain narra-
tive genres, Sigurðsson concludes that Old Icelandic, as well as Modem
Icelandic, was an SVO-language.
Although Sigurðsson’s criticism of Kossuth is well-founded, there
are certain peculiarities in his own study which may make us reluctant
outrightly to accept his rejection of Kossuth’s SV- FS'-hypothesis. Like
Kossuth, Sigurðsson bases his calculations on both main clauses and
subordinate clauses. This might be the correct method if the sole pur-
pose is to find out the basic word order of Old Icelandic, since it seems
clear that L5-order needs certain circumstances. However, if we also
want to determine the stylistic role played by narrative inversion in the
Sagas, it is obvious that we should only count main clauses, since it is
only in such clauses that the writer has the possibility to choose
between different constructions. In his study, Sigurðsson makes the
distinction between main and subordinate clauses for parts of his
material, but since he is not consequent in this respect, there are rea-
sons to believe that his investigation underestimates the role played by
narrative inversion in Old Icelandic Sagas. This warrants another look
at this much discussed construction. In the present paper, I will try to
determine more properly than Kossuth and Sigurðsson the relative fre-
quency of narrative inversion in some Old Icelandic Sagas. In addition,
I will briefly discuss what the result implies with respect to the basic
word order of Old Icelandic.