Íslenzk tunga - 01.01.1961, Side 112

108
HREINN BENEDIKTSSON
(Eyjajjörður) we find the highest proportion, about 96% according
to Guðfinnsson’s calculations. On both sides, the proportion for
harðmœli is lower; in the eastern part of Skagafjörður about 93%,
and in Suður-Þingeyjarsýsla about 89%. Similarly, in the South and
West (from Austur-Skajtafellssýsla to Slrandasýsla), we find for lin-
mœli a percentage ranging from 87% to 97%. In the South and
West, as well as the North, we reach a figure of, or approximating,
100% only if we include speakers with mixed pronunciation.0*'
Above we saw that, for the three early dialect remnants in the
South-East and North-West, the figures for pure pronunciation range
from 36% to 68%, and from 90% to 100% if we include mixed
speakers. For the three areas of flámœli, the highest figure in the
East is about 75%, about 60% in the South-West, and only about
30% in the western part of the North.97
For the remaining two features the same holds true as for linmœli:
In the centre of the North (Eyjafjörður) we find only 74% pure
voiced ð, l, m, n before p, t, k, and 97% pure and mixed; in the centre
of the Southern area (Vestur-Skaftafellssýsla) we find 91% pure
Av-pronunciation, but very nearly 100% pure and mixed.98
The principal characteristic of the dialect situation in Iceland,
therefore, is that, as a rule, for the different features, viz. those that
are losing ground, the mixed areas are much bigger than the pure
ones, and in some cases it is even impossible to speak of any pure
areas at all. A corollary of this is that the boundaries between two
opposed features are vague: In no single case do we find a sharp,
clear-cut boundary between two opposed dialect features. Thus,
whereas the negative concept of ‘dialect difference’ is applicable to
Icelandic, the positive concept of ‘dialect’—as a well-defined, distinct
entity, characterized by certain peculiarities which mark it off clearly
from its neighbours—is less appropriate.
í,,, See Guðfinnsson, Mállýzkur, pp. 206f. The figures for thc caslern part of
Skaga/jörður are calculated on the basis of the table on pp. 181f.
97 See Guðfinnsson, Breytingar, p. 26.
98 Ibid., pp. 19 and 21.