Bibliotheca Arnamagnæana - 01.06.1999, Page 220

200
Conclusion of Part One
in the first instance be taken as an argument for an early date. Noreen also
admits that if a poem has a concentration of archaic features, the burden
of proof lies with those who hold that it is recent.14 On the other hånd, it is
always possible that in an old poem there are no features remaining to in-
dicate an early dating, either by accident, or because its language has
been successfully remodeled during the process of transmission. In no
case I think is there reason to believe that the language of the written text
is identical with the language of the original. Nevertheless, as stated in
the previous part, it seems probable that an old poem will contain a larger
number of archaic features than a younger one; and as I understand it, the
dating consists in nothing but collecting and evaluating such traces.
In my opinion, one important methodological principle follows from
our approach. Since the empirical study aims at the traces within the text
and not the texts themselves, we will adopt an analytical, non-holistic
procedure. Thus we are not going to examine the date of any individual
poem in the first instance, but the validity of individual dating criteria.
This is important in order to avoid having the presence of one criterion
influence the evaluation of another.
Afterwards, it is in principle possible to sum up, and the best one can
expect to find is a convergence of criteria for the dating of some poems,
indicating their place in a relative chronology and some very few hints
of an absolute chronology. That is not the purpose set by the present
work, however. The result of this study of dating can be given only in a
general and relatively abstract form.
A final methodological point is that I have decided not to propose any
new idea whatsoever on how to determine the age of an Eddie poem,
rather to keep to the existing ones, and try to examine their value. The
best reason for this is my conviction that the usefulness of any new
device will always be greatly overestimated by its inventor.
In the following I am thus going to examine a very restricted selection
of dating criteria which seem relevant to me, either because they are li-
able to give empirical results, or because they have played an important
role in the discussion so far.
14 “En faktiskt yngre dikt kan naturligtvis av ett eller annat skal forete vissa ålder-
domligheter, som en aldre dikt kan sakna. [...] Men vi ha redan forut framhållit att dyli-
ka ålderdomligheter nar de sammantraffa i fråga om en och samma dikt naturligtvis
utgora viktiga sannolikhetsskal for hog ålder” (Noreen 1926: 136-37).