Íslenskt mál og almenn málfræði - 01.01.2008, Page 72

70
Ida Larsson
(34) sem er [...] nýbúið að missa [...] ástvin (ístal)
who is new-finished to lose a loved one
‘who has recently lost a loved one’
The account of the restrictions on be in McFadden and Alexiadou
(2006a, 2006b, 2007) relies on a connection between the possibility of
a past counterfactual reading and experiential perfects: when be is not
a temporal auxiliary, it is expected to be impossible in both. The
behaviour of the construction with vera + active participle in Icelandic
corroborates this connection; we have seen that vera is restricted to
the resultative reading, and it is also impossible in past counterfactu-
als. As noted, the construction with vera búinn að is, however, not as
restricted, and it has also often been treated as a perfect (see in partic-
ular Jóhannes Gísli Jónsson 1992 and Wide 2002). In the following,
we consider some contexts where the construction with vera búinn að
has an experiential reading, according to the distinctions in section 1
above. These cases can be accounted for if we assume that vera búinn
að is neither a resultative like vera + active participle, nor a perfect
like the construction with hafa, but that it expresses a resultant state.
The assumption is that a perfect is an expression with double tense
morphology, and (contra Parsons 1990) not necessarily a construction
that denotes a resultant state. In addition, stative passives with be in
languages like German and Swedish can express resultant states (see
e.g. Kratzer 2000, Alexiadou and Anagnostopoulou 2008 and Larsson
forthcoming). As we will see in section 4.1 below, these passives have
much the same connotations as the construction with vera búinn að
has in experiential contexts.
3. Vera búinn að and the experiential perfect
3.1 Introduction
One reason that the construction with vera búinn að has been treated
as a perfect on a par with the hafa-perfect is that that the two ofiten
appear to have the same discourse functions; they express ‘current rel-