Bibliotheca Arnamagnæana - 01.06.1999, Blaðsíða 218

198
Conclusion of Part One
in interpreting the possible analogy, I am doubtful, however, whether
this would lead to more solid results.
The so-called higher criticism of texts treated the poems as compos-
ites, where the separate parts belonged to different periods (for an ex-
ample, cf. NeckeTs analysis of Atlakvida, pp. 122-23 above), and if this
analysis is true, it renders dating extremely difficult, perhaps impos-
sible, since every detail that could serve as evidence for an early date
would be taken out of the actual text in order to be assigned to another,
older poem, partly transmitted by the text in question. Every argument
for an early date would thus tend to slip out of the hånds of the investi-
gator. Fortunately, practitioners of higher criticism have a tendency to
cancel each other out. Their results very seldom coincide, and this faet
greatly contributes to weakening faith in them. They may be right, but
when discussing the question of dating, we most easily avoid begging
the question if we start from the assumption that we have to do with
unified texts, which should be dated as such. If the different arguments
diverge, we may have to consider whether there are additional and
independent arguments for a composite text.
Talking about “reasonable identity” allows for a certain tolerance, but
to what degree I find impossible to define exaetly. I shall give just two
examples, between which the limit of tolerance is to be set. As is well
known, Vgluspå exists in three different versions, none of which appears
to be complete. It is generally assumed that the differences are mainly
due to the hazards of oral transmission.13 Particularly in the order of
stanzas in the middle of the poem there is considerable variance, but the
wording of the common stanzas does not differ so mueh that it seems
unreasonable to try to date it as one text. On the other hånd, the Nor-
wegian ballad Torekallsvisa recounts parts of the same story as Pryms-
kvida\ but it is obviously another text - the language as well as the genre
are different.
Erik Noreen argued that it is possible and even probable that some
Eddie texts are older than the Old West Norse language in which they
are transmitted, i.e. that they were composed in the Proto-Norse period
(before 700), and that some of them are translations of South German
texts (Noreen 1926: 130-32; cf. chapter X below, p. 262 ff). When try-
ing to date a text by inter alia linguistic and metric means, it is tempting
13 For another view, cf. 6lafur M. Olafsson 1967: 110.