Gripla - 2023, Page 23

S I Ð R , RELIGION AND MORALITY 21
of siðr to understand Viking Age thought, even though the word only
survives in texts from the last years of that period? And why does siðr not
appear in earlier texts, regardless of its definition? Neither of these have
firm answers; the only response can be hypotheses based on later use and
current understandings of early northern cultural trends.
Where is the Viking Age siðr?
The scarcity of siðr in early texts, regardless of meaning, could be an acci-
dent of preservation or signal that the terminology did not gain importance
until later. The former is not a radical suggestion, given the low quantity of
poetry that has been preserved overall and of instances of siðr within that
body, as well as the environment in which Old Norse texts were transmit-
ted, in which material addressing non-Christian religion directly is less
likely to have survived (Taggart 2021, 286–87). On the other hand, while
kennings based on Old Norse mythic tropes fall out of use in the eleventh
century (Clunies Ross 2005; Males 2017), some poetry utilizing them does
continue to be transmitted. It would seem quite an accident for a compara-
tively neutral term like siðr to be wiped out when those kennings were not.
Attempting to increase the clarity of this picture, I have counted in-
stances of siðr (simplex or in compounds but not ósiðr) alongside stanzas
and fragments of verse that survive from the ninth century until the elev-
enth (Figure 1), reckoning each stanza and fragment as a unit regardless
of length and using the dates given by Skj. (BI). Unfortunately, this infor-
mation can only provide a suggestion of the past reality. A stanza of ten
lines has the same weight in these calculations as a fragment with two, and
Finnur Jónsson’s datings can be queried on the basis that poetry may be
inauthentic and that a poet is counted in a single century even when their
work spans two (arguably the year 1000 is the only boundary meaningful
for its own sake, due to the Alþing’s conversion). However, Figure 1 would
not be much more dependable even if the dates were painstakingly scru-
tinized, eddic poems included, and individual lines counted, given that no
one knows how many verses have been lost from each century (particularly
from non-Icelanders). Likely, proportionately more poetry is missing for
every century counted back in time. A total of the poetic units that were
actually composed in each century might articulate a very different trend
than the quantities surrendered by today’s fragmentary corpus.