Sequence of Tense and the Speaker’s Point of View: Evidence from the Imperfect

The idea that the imperfect has a basically quantificational meaning has the consequence that the so-called continuous readings, as in Alle cinque Mario dormiva (at five Mario slept(IMPF)), tend to be either disregarded or explained as instances of the partitive phenomenon (Krifka 1998). In previous work (Giorgi and Pianesi 2001a; see also §2 below) we argued that the partitive analysis of the imperfect (and, generally, of the imperfective aspect) is inadequate, both empirically and theoretically. We believe that in order to understand the properties of the imperfect tense in Italian – and in Romance languages in general – the continuous readings, and the many others that do not seem to directly involve quantificational phenomena need be taken as prime source of evidence. This can be accomplished by acknowledging a basic distinction in the domain of eventualities between terminated and non-terminated ones, by readdressing the telic/atelic divide, and by rethinking imperfectivity/perfectivity as a morpho-syntactic phenomenon, and separating it from its notional counterpart. In this paper, we will review, and provide details about, some of these points. We will propose a view of the Italian imperfect which crucially relies on two intuitions (and which, taken separately, are by no means new): a) that in some sense to be made more precise,