Íslenskt mál og almenn málfræði - 01.01.1980, Qupperneq 180

178
Joan Maling
counted as subjects just in case the finite verb comes next. Thus, in the
word order study referred to above, Kossuth states that „the relative
particles sem and er were counted as C( = conjunction), but if appro-
priate, also as S( = subject)“ (Kossuth 1978:446). It is clear that what
is intended by „if appropriate“ are sentences in which the subject NP
has been relativized, and hence there is no subject to count as filling
first position. Similarly, in his discussion of the possible positions of the
finite verb in Scandinavian languages, Haugen gives the (rather stilted)
example shown in (3a) with the indicated numbering:
1 2 3 2 3
(3) a Þeir vissu, hver genginn væri/væri genginn (Haugen(1976:84))
1 2 3 2 3
b Maðurinn, sem farinn var/var farinn heim, heitir Pétur
Haugen does not give an example of a comparable relative clause, but
he would presumably count the position of the finite verb as indicated
in (3b). Note that the finite verb is counted as being in third position if
the past participle has been inverted.
It is precisely for these cases that I wish to propose a change from
the traditional treatment of word order, namely, that even in these
cases, the finite verb should be counted as being in second position, as
illustrated in (4):
1 2 12 3
(4) a Þeir vissu, hver farinn var/-var farinn heim
1 2 12 3
b Maðurinn, sem farinn var/------var farinn heim, heitir Pétur
1 2
c Maðurinn, sem--------kom í gær, heitir Pétur
If the past participle is inverted, then it counts as filling first position;
it it is not inverted, then the first position of the embedded clause will
be considered empty, as indicated by the dash. In either case, the finite
verb is in second position. The same is true of an embedded clause with
a simple past tense verb as given in (4c): the subject NP has been rela-
tivized, leaving first position in the relative clause empty, and the finite
verb kom is in second position. The relative particle sem is not counted
as subject.
What I am suggesting, then, is that sentences such as those in (4)
are simply another type of subjectless sentence. There are two advan-