Íslenskt mál og almenn málfræði - 01.01.1982, Blaðsíða 117

Um tvenns konar It-framburð
115
/(-pronunciation has never been investigated systematically. But it seems to be strongest
in Eyjafjöröur (particularly in Svarfaðardalur) near the western limits of the „voiced
area“.
In 1.2, references are made to earlier scholars. The first to observe and make clear
that voiced / was not used accidentally but according to fairly well established rules
was Stefán Einarsson 1928-1929. Sincc thcn, practically nothing has been added to
our knowledge about this dialect feature. No „native speaker" of this dialect has contri-
buted to the literature on Icelandic dialectology until now.
In Chapter 2, the rules of voiced and voiceless / before /t/are presented and discussed.
Stefán Einarsson's rules are in most cases confirmed, but some corrections are made
and new material addcd. The rules are enumerated in 2.5. In Chaptcr 5, they are
summarized as follows:
/ is voiceless when the spelling is ‘llt’. In the cluster ‘lt’, thc / is voiceless if
both sounds belong to the same stem morpheme (bylta) but voiced if thc / is
immediatcly followed by a morpheme boundary (gul+t from gulur). Thc main
exceptions are the verb forms skalt, vilt (2nd pers. sing. in pres.) and unstressed
syllables, in particular suffixes (gamall), — in these forms thc / is voiceless.
Finally, / is voiced before/t/ if an intermediate sound is droppcd (álft).
In other words, the voicing (and voicelessness) of / beforc /t/ is, to a great extent,
morphologically conditioned (gult), in some cascs phonologically (álft) and evcn lcx-
ically (púlt).
As illustrated by examples in Chapter 3, there is a distinctive contrast betwcen voiced
and voiceless / when followed by a /t/ (mœlt (pret. mœldi) ~ mœlt (prct. mœlti)). Fur-
thcrmore, it is shown that both It's are distinguished from Id (holt (adj.) ~ liollt ~hold).
In Chapter 4, some considerations arc presented on the age and origin of voiced
vs. voiccless / bcforc /t/ witli respect to the two types of /, cacuminal and dental, having
been suggested for Old Icelandic, and thc two types of / before /d/, evidenced by spel-
lings as well as rhymes in Icelandic from the 14th century until about 1700.
Háskóla íslands,
Reykjavík