Íslenskt mál og almenn málfræði - 01.01.1985, Síða 134

132
Christer Platzack
more language dependent, naturally, but it is not too difficult to make
the adjustments that makes it a valuable tool for the analysis of Old
Icelandic. Cf. Platzack (1984), where these adjustments are discussed
in some detail.
When doing quantitative studies of a certain language, it is impor-
tant not only to have a reliable way to analyze the language, we must
also have an understanding of to what extent different principles are at
work in different parts of the grammar. In my opinion, Kossuth and
Sigurðsson both fail in this respect when they base their frequency
figures for the narrative inversion construction on the total number of
clauses, main clauses as well as subordinate clauses. As Sigurðsson
himself has shown, narrative inversion is a root phenomenon in
Emonds’ terms (Emonds 1976); i.e., it is a phenomenon reserved for
main clauses. If we are interested in finding out how frequent this word
order pattem was in Old Icelandic Sagas, i.e., if we want to say some-
thing about the style of these texts, it must be misleading to give the
frequency of FS'-clauses as a percentage of the total number of main
and subordinate clauses — only declarative main clauses would do,
i.e., declarative macrosyntagms in the terms of Loman & Jörgensen
(1971).
The decision to keep main clauses and subordinate clauses apart
when it comes to the establishing of word order pattems in Germanic
languages is also theoretically well founded. As Diderichsen (1941:58
f.) concludes, the principles determining the placement of the finite
verb are different in main clauses and in subordinate clauses. As has
been shown in several modem studies of Germanic word order, the
position of the finite verb in main clauses principally equals the posi-
tion of the complementizer in subordinate clauses. Cf. studies by den
Besten (1983), Holmberg (1983a), Platzack (1986) and several others.
Thus, e.g., there is a subject position immediately behind the finite
verb in main clauses, and immediately behind the complementizer in
subordinate clauses.
(3)a Ekki munu vit svá skjótt skilja ... (Grettis saga, 1937:176)
b þá kvað Grettir vísu (Grettis saga, 1937:177)
c Tekr bóndi við honum vel (Grettis saga, 1937:110)