Íslenskt mál og almenn málfræði - 01.01.2008, Side 86

84
Ida Larsson
In Modem Swedish, stative passives like tliose in (56) and (57) are
often contextually restricted; they typically occur e.g. in lists of things
that have been done. Particularly when the participial verb is atelic,
the event needs to be planned and expected, and the passives tend to
be better when expressing e.g. every-day household tasks than other
things. While none of the examples with anvánda ‘use’ in (57) are
marked, examples like those in (59) below are. As pointed out by
Kratzer (2000) with regard to the corresponding German passives, the
examples in (59) sound “bizarre out of the blue, but as soon as we
impose a ‘job is done’ or ‘that’s over’ interpretation they become
fíne” (2000:4). For (59a), we can, for example, imagine a situation
where somebody has to take care of a cat and make sure to pet it once
a day. Hence, the implications are similar to those of the examples
with the construction with vera búinn að discussed in section 3 above:
(59)a. Katten ár redan klappad.
the cat is already petted
‘The cat has already been petted.’
b. Bamvagnen ár redan knuffad.
the baby-carriage is already pushed
‘This baby carriage has already been pushed.’ (Cf. Kratzer 2000:4)
Since the stative passives in the examples in (57) and (59) involve
atelic verbs, they clearly cannot express target states as the resultative
with vera + active participle in Icelandic arguably does. Kratzer
(2000) suggests that they instead denote resultant states, i.e., unspeci-
fíed ever-lasting states following from a previous event. Following
Parsons (1990), who introduced resultant states to account for the
semantics of the perfect, Kratzer assumes that resultant state passives
express perfect semantics. Under the present account, this can clearly
not be maintained. In Larsson (forthcoming), I argue that the partici-
ple in resultant state passives has perfective aspect but lacks the tense
specification of a perfect.19 In fact, a ‘job is done’ reading is often
19 A different possibility would be to assume that “resultant state” is an aspectu-
al value on a par with the imperfective and the perfective; this has been suggested by
Ramchand (2008).