82
problem in the classical kenning system: while
it is often assumed that the referent of
SPEAR
exists in the system as a separate and
distinguishable entity, in reality it is poorly
distinguished by formal means alone. It does
not seem to have developed a kenning-model
that is stable and unique to it, which is
necessary to prove any referent’s existence.
Meissner (1921: 145) says this explicitly:
“Kenningar des Speers sind [...] von den
Schwertkenningar nicht immer mit Sicherheit
zu sondern” [‘It is not always possible to
reliably distinguish between the kennings for
SPEAR
and those for
SWORD
’]. The material
indeed shows that some basewords for alleged
kennings for
SPEAR
are identical to those used
for kennings of
SWORD
, while other basewords
that seem to denote either a
SPEAR
or an
ARROW
are in fact identical to typical
BAD WEATHER
basewords for kennings of
BATTLE
(e.g.
él
[‘snow-shower’]),
further complicating the
distinction. This is because even in the case in
which such a baseword is used with a
determinant that is a heiti for
PART OF BOW
(resulting in a hypothetical model ‘projectile of
bow-part’, where ‘projectile’ prime is
extracted from a ‘bad weather’ word, where
e.g. rain droplets are ‘seen’ as projectiles, and
the model meaning
ARROW
), this heiti can be
understood as simply a heiti for
BOW
. Thus, in
the case of a suspected
ARROW
kenning, a
determinant that ‘wants’ to be recognized as
PART OF BOW
, will nevertheless be recognized
by default as simply a
WEAPON
determinant,
and the detected model will be ‘bad weather of
weapon’, i.e. a core kenning-model for
BATTLE
, not the hypotheticial ‘projectile’
model for
ARROW
.
This
means that any kenning that
intends to
be a kenning for
SPEAR
is always in danger of
being misidentified as either one for
SWORD
or
for
BATTLE
, and the system of kennings has no
internal way out of this conundrum, not having
developed a list of basewords and determinants
from which the combination would yield a
uniquely recognizable kenning-model for it.
And our poet is clearly aware of this,
attempting a
system-external way (i.e. one not
relying exclusively on baseword/determinant
variation) to resolve the ambiguity.
Helgi Harðbeinsson’s spear is mentioned
twice in stanza 7, each time referring to it with –
necessarily, as we have just discussed – a
SWORD
kenning: the first is the above-
mentioned
hring-þollr (itself mistakable for
the kenning of
MAN
/
WARRIOR
), the second is
sára-teinn [‘tooth of wounds’]. And to the
latter he attaches an epithet, an agreed adjective
breiðr [‘broad’] (7.3
sáratein breiðan). This is
remarkable because this epithet is specifically
used for spears, e.g. in the famous words of
Grettir’s brother Atli in
Grettis saga 45 as he
is being killed with a broad-tipped spear:
Þau
tíðkast nú, in breiðu spjótin [‘They are quite in
vogue these days, those broad-tipped spears’].
What is also remarkable is that the reason this
particular adjective is picked is purely formal,
that is, the adjective is picked with the specific
aim of solving the problem of kenning-system
ambiguity, and its choice is
not driven by the
actual words of
Laxdæla, which in fact it
contradicts. While the spear that killed Atli in
Grettis saga is indeed broad, the one that Helgi
Harðbeinsson kills Bolli with is specifically
described as ‘long’:
Helgi Harðbeinsson ok
hafði í hendi spjót þat, er alnar var lǫng fjǫðrin
[‘And now Helgi son of Hardbein held a spear
in his hands, one such that had a blade a whole
ell long’] (
Laxdæla 55). Had our poet picked
the saga word, his attempt to solve the
ambiguity would have failed, as ‘long’ would
be assumed to reinforce the reading of the
kenning in question as
SWORD
. By picking
breiðr, the poet provides the reader/hearer with
a trigger to overrule the basic kenning pattern
recognition
rule,
proving
his
in-depth
awareness both of a particular problem within
the traditional system and of the system-
compliant way this problem might be bypassed.
Features Retained from Classical dróttkvætt
As we noted above, despite the features
discussed thus far, the poet’s attempt to imitate
classical
dróttkvætt is largely successful. This
is due to the retention of many key features,
besides the reproduction of the obvious bulk of
metrical rules regarding alliteration, stem-
rhyme, use of the close, pairing of odd and
even lines, syllable count (even if with 18
th
-
century syllables), prosodic structure of the
fillers of the various positions and so forth.
One particularly famous quirk of classical
dróttkvætt is its split-sentence technique
(sometimes called ‘intercalation’): in the flow
of the verse, sentence A can suddenly be