Íslenzk tunga - 01.01.1961, Qupperneq 115
ICELANDIC DIALECTOLOGY: METHODS AND RESULTS
111
tion, especially by Norwegian scholars,101 of the existence of two
distinctive word tones (acute and grave) in Icelandic. What has
happened is that to certain observable phonetic phenomena in Ice-
landic these scholars have ascribed a linguistic function which
similar phenomena have in their own language system, hut which is
absent in Icelandic.
From the functional point of view, we see that some of the existing
differences consist in positional variation of the phonemes (the
voiced vs. voiceless pronunciation of ð, l, m, n before p, t, k),
especially in positions of neutralization (harðmœli vs. linmœli).
Others pertain to phonemic distribution or incidence in the voca-
bulary (e. g. monophthongs before [q] in the North-West, and before
[j] in the South-East). But, as a rule, the same significant distinc-
tions are made in all varieties of the language, and the phoneme
inventory is practically the same. The most notable exception is the
existence, in the South as opposed to the North, of the voiceless velar
spirant, labialized or not, as a separate unit, spelled hv-.102 Thus,
101 See, e. g., Hægstad, Nokre ord, pp. 3—4.
102 Also in flámœli a loss of contrast is involved, when i and c, and u and ö,
are confused. This merger is most widespread in the long variants of these vo-
wels, but it is to be found in the short variants, too (Guðfinnsson, Mádlýzkur,
pp. 56 and 113). Possibly, therefore, there are varieties of Icelandic with a
reduced vowel phoneme inventory (see “The Vowel System of Icelandic_____,”
pp. 305f.).
Other dialect differences of contrast are: (1) In the North-West, with
monophthongal pronunciation before [IJ], the contrast which exists elsewhere
between, e. g., hengdi ‘hanged’ lhdl,l(]i | and hegndi ‘punished’ theij cjtl is
prohably ahsent (see the present writer, “Nokkur dæmi um áhrifsbreytingar í
íslenzku,” Lingua Islandica—íslenzk tunga I (1959), p. 60); (2) In the Soutli-
East, with monophthongal pronunciation before [jl, there is contrast between,
e. g., lagi ‘order’ (dat.) Ila:ji] and lcegi ‘anchorage’ [lairjl], legi iiquid’ (dat.)
[le:ji] and leigi ‘(I) rent’ [lei:jll, lögin ‘the law’ [lö:jin] and laugin ‘the hot
spring’ [löy:jin], stigi ‘(I) stepped’ (pret. subj.) [sqji:ji] and stígi ‘(I) step’
(pres. subj.) [s^ji:ji], which are, two and two, homophonous elsewhere (llaiji,
leiji, löyjin, s<J(i)iji]); (3) In the North we find, sporadically, a three-way con-
trast, e. g., mœlt ‘spoken’ (past part. neut.) fmai] vs. mœlt ‘measured’ (past