Studia Islandica - 01.07.1963, Qupperneq 99

97
A more differentiated picture of the frequency, however, will be
obtained by studying the distribution of the 35 words over smaller text-
portions, sections of 5000 words, as is demonstrated in Table 6 on p. 34.
(The concluding figures in parenthesis for each text denote an incom-
plete 5000-word section.) It is a salient feature of this table that the
columns for Laxdœla and Knýtlinga appear to one’s eye as “thicker”
than those for the control material. The highest figure for the latter
is 16; it appears twice — in Oddr — in the whole bulk of 102 sec-
tions. In Laxdœla and Knýtlinga only 3 out of 20 sections are beneath
that frequency, and only one section has a lower number than 10. In
the control material, on the other hand, as many as 79 out of 102
sections fall beneath 10.
It is an interesting fact that the two highest figures, 57 and 43,
appear in Knýtlinga — and not in Laxdœla, as one would have imag-
ined. The fact becomes still more interesting on a closer examination
of the sections in question, the sixth and the seventh. For these sec-
tions cif the saga, scholars have not been able to find any apparent
prose sources. Of the seventh section Gustav Albeck remarks that its
structure has an epic touch (“er sagamœssig i sin Opbygning”), and
that it reveals “much independence”. This seems to indicate that the
author had comparatively a free hand in composing those parts of his
story — a situation which apparently favoured his individual phraseo-
logy.
In order to test the result of the foregoing comparison, more con-
trol material, both family sagas and Kings’ sagas, was examined, in
all 609000 words distributed over twenty-five various texts (pp. 36
—37). The frequency of the adjectives and abstract nouns in question
is shown in Tables 7 (p. 38) and 8 (p. 39) respectively. For both series,
the quotients of each separate text were calculated in the simple way
described above. On pp. 40—41 the numbers from the two series are
added together.
The new control material reveals throughout somewhat higher fig-
ures than the original one; its average quotients for the adjectives
and the nouns are 1.4 and 1.6 respectively. But this should not puzzle
us, since the high-frequency words in Laxdœla were picked out in re-
lation to the original seven control texts.
Summing up the result, one can state that neither in the adjective
series nor in the noun series do any of the thirty-two texts in the control
material, comprising more than a million words (529000 + 609000 =
1138000), reach the frequency for Laxdœla (3.9) or Knýtlinga (3.2).
The quotient which comes next, that of the short Bandamanna saga,
is 2.6. Thus Laxdœla and Knýtlinga in their choice of words reveal
a clear and exclusive affinity. Otherwise they have no real points of
7
L