Bibliotheca Arnamagnæana - 01.06.1999, Side 295

XI Mythological kennings
275
in Iceland from the early post-conversion period, e.g. stanzas in Banda-
manna saga and in HofgarSa-Refr’s poem on Gizurr gullbrårskåld,
mythological kennings do not seem to have been avoided there. In the
case of Arnorr jarlaskåld it can even be shown, according to Kuhn, that
he was less scrupulous in avoiding heathen allusions in his poems to the
earls of Orkney than in his Norwegian court poetry. In his 21 Orkney
stanzas there are no less than 6 kennings alluding to the heathen religion
(in addition to his famous Ragnargk-allusion in Porfmnsdrdpa st. 24),
whereas 48 stanzas in poems for the Norwegian kings contain only 4
such allusions.
This last example illustrates the statistical problems implied in
Kuhn’s reasoning. However justified his claim may be that the material
should be subdivided according to nationality, it leaves us with such tiny
figures that no reasonably safe conclusion can be drawn from them.
Fourteen years later de Vries replied to Kuhn’s criticism in the same
periodical with an article called “Kenningen und Christentum” (1956).
He paid tribute to Kuhn’s claim that the data ought to be split into
groups according to origin, but he questioned whether this would have
any impact on the results, given the small amount of Kuhn’s counter
evidence. He then proceeded to discuss this in some detail, trying to do
away with the little evidence Kuhn had been able to accumulate. Kuhn
had mentioned four allusions to heathen myths in the poetry of one of the
few sagas believed to quote stanzas from the period after 1050, Banda-
manna saga (Kuhn 1942: 141 = 1971a: 303), but de Vries objected
that Gu5ni Jonsson had surmised that the stanzas in question were later
interpolations, which could not be taken as valid examples of poetry
from this period.12 In the case of Arnorr jarlaskald, de Vries tried to ex-
plain away one of the mythological kennings by the faet that an instance
of the same kenning is found earlier in a poem by borbr Særeksson, who
in several cases seems to have been imitated by Arnorr. At most, this
heathen kenning can be taken as an instance of the uncertainty of a be-
ginner, de Vries maintained (1956: 128). This is a rather dubious line of
argument, however. If we are entitled to discount, or attach less import-
ance to, cases that can be taken as an imitation of older exemplars, this
12 De Vries 1956: 126. De Vries here refers to GuSni Jonsson’s introduction to Islenzk
Fornrit 7, which appeared in 1936, later than his own treatise from 1934, where he had ac-
cepted the evidence from Bandamanna saga for this period without discussion.