Studia Islandica - 01.07.1963, Side 102

100
Islandica 20, pp. 14—23, and pp. 182—84 (Summary). For fuller in-
formation conceming this method, the reader is referred to that paper.
Here a rough sketch must suffice to supply the necessary background.
In this connection a “pair word” means a word which in the
whole bulk of the texts occurs in Knýtlinga, on the one hand, and
in one only of the five family sagas consulted (Egla, Laxdœla, Eyr-
byggja, Njála and Grettla), on the other. The frequency of such pair
words is interpreted as a measure of the affinity between Knýtlinga
and the saga in question.
For these statistics the entire vocabulary was not examined. AIl
nouns for concrete things were omitted, as they are mainly deter-
mined by the subject matter and not by the author’s individual mode
of expression. The remaining vocabulary was divided into four diffe-
rent categories: I. Adjectives, both concrete and abstract. II. Nouns
denoting persons or groups of persons. III. Verbs or verb construc-
tions. IV. Abstract nouns (in a broad sense).
In the Snorri-Egla investigation Heimskringla played the part now
played by Knýtlinga. As Heimskringla is not engaged in the present
inquiry, a word is of course considered as a pair word hetween Knýt-
linga and one of the five family sagas, though it may happen to occur
in Heimskringla too. But for certain reasons, which will be accounted
for presently, it has been found convenient to treat the “Snorri words”
separately. In Appendices I—IV, which record all the pair words for
each category and each saga, the words that are not also found in
Heimskringla are marked with an asterisk. Thus in Appendix I:a the
adjective bjúgr ‘crooked, curved’ occurs once (p. 54) in Knýtlinga,
once (p. 277) in Egla, and twice (II 187 and III 232) in Heims-
kringla. The epithet gullsaumaSr ‘embroidered with gold’, on the other
hand, is not to be found in Heimskringla, only once in Knýtlinga and
Egla, pp. 208 and 213 respectively. (Knýtlinga is cited from the edi-
tion by Carl af Petersens & Emil Olson: S$gur Danakonunga, Koben-
havn 1919—1925, the other texts from the editions in the series ls-
lenzk fomrit, Reykjavik 1933 etc. The Roman figures I, II and III
in connection with Heimskringla refer to volumes in the Islenzk fom-
rit edition. From Heimskringla only three instances of each word are
referred to; if there should be more of them, this is indicated by “etc”.)
The numbers of pair words for each category and saga are given
on p. 53. But as the texts compared differ considerably in size, a simple
summing-up does not give a correct picture of the relative frequency.
We need some standard measure to which we may adjust the figures.
Of course any of the sagas, or any arbitrary size, could have served
as standard. But as our interest is now focused on Laxdœla it was
found convenient to choose that saga with its 58000 words for the