Íslenskt mál og almenn málfræði - 01.01.2008, Page 89

Becoming Perfect: Observations on Icelandic vera búinn að 87
gy. In a similar way, several separate elements presumably contribute
to the temporal-aspectual properties of the construction with vera
búinn að: the tense and aspect of the stative auxiliary vera, the resul-
tant state semantics of the participle búinn, and the aspect of the
embedded predicate.
Imperfective aspect can be assumed to focus only part of an event,
without making reference to its endpoint (see e.g. Smith 1991). When
the embedded predicate has imperfective aspect, the construction with
vera búinn að will therefore express that a part of an event precedes
the reference time, without asserting that the event has been complet-
ed or flnished; with a durative adverbial, it can assert that the event
has not come to an end. Since a resultant state can follow from an atel-
ic event like e.g. ‘sitting’, nothing excludes that it can also follow
from a subpart of it. In fact, atelicity is generally defrned in terms of
homogeneity: an event of sitting is atelic since any subpart of it is also
an event of sitting (see e.g. Krifka 1998).
I leave the details of the interaction between the elements involved
in the construction with vera búinn að for future work; what is impor-
tant here is that an account of vera búinn að as a resultant state con-
struction does not, in principle, exclude the universal reading.
5. Conclusion
In this paper, I have argued that the modem Icelandic constmction
with vera búinn að has not yet developed all the properties of a per-
fect tense, despite the fact that it has a meaning very similar to that of
a perfect. By assuming that vera búinn að expresses a resultant state
without the tense morphology of the perfect, we can account for the
fact that vera búinn að is impossible in past counterfactuals, as well as
ihe fact that búinn sometimes retains lexical meaning. The assumption
a>so explains the “resultative flavour” of many examples with vera
búinn að as observed by Jóhannes Gísli Jónsson (1992) and it
accounts for the semantic similarity to resultant state passives in lan-
guages like Swedish. I have suggested that the temporal-aspectual