Íslenzk tunga - 01.01.1961, Qupperneq 104
100
HREINN BENEDIKTSSON
For flámœli—the merger of i and e, and of u and ö—the situation
is different. This is the only case where an innovation is not to be
found in one continuous area, but occurs, instead, in three separate,
isolated areas, the East, the South-West, and the western part of the
North. There are no cogent reasons for assuming that this feature
spread from one of these areas to the others. Therefore the opinion
has been expressed that it originated independently, by parallel
development, in at least two of these areas, the East and the South-
West.70
As regards the earlier innovations, their areas of origin and the
direction of their extension are much more uncertain. For the diph-
thongizations of original long é and œ we saw that Bishop Brynjólf-
ur Sveinsson and Árni Magnússon were of the opinion that they had
70 Dahlstedt, “fslenzk mállýzkulandafrœði,” pp. 40f. Dahlstedt is of the
opinion thal in the East, where ancient long open œ was not diphthongized to
Tai], the immanent structure of the vowel system naturally led to the merger
implicit in jlámœli. Where œ was not diphthongized, the front unround vowel
series, consisting of [e] — Te] — [1] — [i], exemplified by væður ‘wadable’ —
veður ‘weather’ — viður ‘wood’ — víður ‘wide’, was overloaded, and thus pro-
vided the most favourable conditions for the merger. However, it is more likely
that when the quantity correlation in vowels disappeared, probahly not later
than the sixteenth century, early long open œ, where it had not become a diph-
thong, merged with original short e (see the present writer, “The Vowel System
of Icelandic: A Survey of Its History,” Word XV (1959), pp. 300f.). At least,
iti those examples where the monophthong is still preserved, the vowel is identi-
cal with original short e. Besides, the alleged systemic conditions account only
for the change in the unround vowels, not for the merger of round u and ö. Also
for chronological reasons, jlámæli, which is very recent, can hardly be related to
the retention of monophthongal æ.
In support of his explanation Dahlstedt (p. 58, n. 24) adduces the close
pronunciation, noted by M. Hægstad in 1907 (see Nokre ord um nyislandsken
(Oslo 1942), p. 44), of the present long variant of original short e, as, e. g., in
tekur. This, however, is undoubtedly not a remnant of an earlier close pronuncia-
tion in contrast with the open pronunciation of æ. It is rather a consequence of
jlámœli: the vowel resulting from the merger of i and e is very often of an inter-
mediate quality, mono- or diplithongal, [e] or [te], and therefore closer than e
elsewhere.