Bibliotheca Arnamagnæana - 01.06.1999, Blaðsíða 222

202
Conclusion of Part One
is futile to try to detect and date the linguistic forms of generations of
oral poets beneath the work of the final improviser. Aldo Ricci [1929:
261] assumes that “the Charms underwent centuries of oral transmis-
sion” before they were first recorded, but the linguistic tests make no
claim to detect that sort of pre-history in a text (Amos 1980: 10).
Given that the written history of Old Norse poetry is much shorter than
that of Old English poetry, applying this sagacious restraint to the dating
of Eddie poetry would amount to backing away from the task. I there-
fore assume that it is possible that different chronological layers from
the poems’ oral (pre-)history may have left sufficient traces in the writ-
ten versions to be detected by a judicial scrutiny of the texts.
Another problem is the intrinsic difficulty in assigning any particular
date to different words or word forms from the preliterary period. Runic
inscriptions may give some points of reference, but for several reasons
skaldic poetry is far more important. In some respects the occurrence of
a word in skaldic poetry may be an even better source than a written re-
cord due to the complicated metrical rules, which will in many cases
show for example the quantity of a vowel.
Admittedly, it is perfectly possible that such trust both in the stability
and in the datability of skaldic poetry is exaggerated. In this context,
however, I adopt the basic assumption that the linguistic evidence taken
from skaldic poetry is in general to be trusted if there are not particularly
strong grounds for doubting it. Apart from the well-known general argu-
ments in favour of this position, I want to underline its heuristic value,
since a hypercritical attitude towards skaldic poetry would deprive us of
our most important point of comparison for the language of Eddie
poetry. In as much as the full truth of the matter is unknown, we are
justified in not being excessively critical. Instead we must make an ef-
fort to be as consistent and systematic as possible so as to tum to
account the evidence we actually have.
The most interesting attempt at dating Eddie poetry on the basis of the
skaldic history of one particular word is in my opinion Hans Kuhn’s hy-
pothesis that the frequency of the expletive particle of/um in skaldic lan-
guage may have some bearing on the dating of a poem. This hypothesis
will be discussed more extensively in the following chapter. The discus-
sion will also serve as an example of the abstraetness to be expected
from the answers to our questions (cf. p. 200 above).