Rit (Vísindafélag Íslendinga) - 01.06.1970, Blaðsíða 465

463
that there is a classificatory distinction of [-voiced] in the
segments underlying [p, t, k, f, s] vs. [ + voiced] in those
underlying the rest, and that aspiration and voicelessness in
the stops (and in fricatives following aspirated stops) is in-
troduced by rules. This may be little more than a question
of terminology.
In weak position there is only one stop series, mostly
without aspiration and thus phonetically best rendered as
[b, d, g]. In word-final position the stops may be aspirated
(and the coronal stop may exhibit the affrication typical of
Danish [t]). In this paper final -t is given as [t] for convenience,
although [d] is probably heard more often.2
The sparseness of stops in weak position is compensated
for by the occurrence of two additional fricatives, [ð, y], and
of two semivowels, [i, u]; the sound [i] is, of course, phonet-
ically closely akin to the [j] of the strong positions.
According to the interpretation in terms of tenseness the
classificatory feature distinguishing [b, d, g, f, s] must be
[ + tense] or strong; the other segments are [-tense] or weak.
In the terminology used here the segments underlying at least
some occurrences of [b, d, g, f] and underlying [s] are [-voic-
ed], those underlying the rest (and underlying some occur-
rences of [b, (d), g, f]) are [ +voiced]. (This is in good agree-
ment with the phonetic manifestation, although [b, d, g] may
become voiced in intervocalic position in rapid speech.)
2. Matching of Initial and Final Consonants
The matching of initial [p, t, k, b, d, g] with final [b, d, g,
2Hansen (1956:50) and Diderichsen (1957:47) rightly consider the optional
aspiration of word-final stops as being related to a general tendency toward word-
final aspiration (before pause) in Danish. From this point of view it is more mea-
ningful to transcribe [-b, -d, -g] than [-p, -t, -k], the optional aspiration being a
junctural phenomenon. One might, on the other hand, question whether it is
phonetically reasonable to transcribe the coronal stop as [d] in cases where it is
clearly affricated. (Admittedly, this phenomenon may also be confined to the posi-
tion before a pause.)