Editiones Arnamagnæanæ. Series A - 01.10.2003, Blaðsíða 36

18*
General introduction
[þær er] þeir skarv hvalinn með’ (They were quite without weapons except
the axes with which they were cutting up the whale.) Actually, this is the
probable reconstruction of the sentence, since at first glance the MS reads
‘eifat’. But the second letter is in fact <y), not <i>; though faint, the letter
can be seen to have, or have had, a tail and also a dot above it, which is
customary for <y> in the handwriting of the period; moreover the word is
paler than the surrounding writing and seems to have been erased. (I am
grateful to the late Jón Helgason who, when he was General Editor of the
series and my supervisor, pointed this out.) The word eyfit (eyvit) is not
very common in mediaeval Icelandic; it is glossed by Fritzner “. . . med
samme Betydning som ekki vætta” (i e “not at all”). Perhaps it was not
understood and the text was altered to ‘fátt vápna’ (“few weapons”), which
is rough sense, though not very good sense, whereas “no weapons at all”
makes excellent sense. In short, E’s is the lectio difficilior. If this inter-
pretation is correct, the common error of ‘ei fat’ (if error it be) implies that
W and *A have a common ancestor, possibly a sister of E rather than E
itself. This derivation, is, however, incompatible with the stemma on p. 17*.
The other apparently unassailable locus occurs in M. That manuscript
has a number of errors of transcription, especially haplography, so as it
stands it can hardly be an early, authentic text. But it is not unreasonable to
assume that it is a copy, or a copy of a copy, of an early version. For there
is one lexical item that seems particularly early, the word gyss m. (‘gys æ
kafliga. heitur’ M 26.50), about which Davíð Erlingsson44 has written to the
effect that while ‘gush’ might very reasonably be regarded as the original
derivation of the word, it lost that meaning relatively early and took on the
meaning of ‘joke’. The noun refers to the sauna bath Styrr had prepared. All
other available independent MSS read hus - ‘varð þat hus ákafliga heitt’, a
possible but not entirely satisfactory reading; at the same time a transition
of ‘gy(s)’ to ‘hu(s)’ is a quite reasonable scribal alteration. The acceptance
of this argument would imply that all other MSS - with the possible
exception of E - had a common ancestor.
I am unable to explain this crux other than by the suggestion that there
has been contamination in at least one manuscript through a scribe’s having
used more than one exemplar; this might happen because his main
exemplar was illegible in places. M, for example, may have had more than
one source; the first three chapters seem to be copied from another manu-
script than that used for chapter 4 onwards (see below, p. 75*). This does not
solve the immediate problem, but does show that a scribe could have access
to more than one exemplar.
44 Davíð Erlingsson 1987.