Overt and covert modality in generic sentences

International audience1. Aim of the paper This paper is about the modal interpretation of Indefinite Generic Sentences (ISs), illustrated in (1). (1) A lion has a mane On the assumption that ISs involve a silent generic quantifier GEN, which is taken to be a modal operator, it has been argued (Krifka et al. 1995) that GEN can have different interpretations, corresponding to those of overt modalities in ISs sentences. In this view, (2-a) would be interpreted as in (2-b), and the silent operator GEN in (2-a) would have the deontic flavor of the overt modal must in (2-b). One of the aims of this paper is to show that GEN is not the silent counterpart of overt modality. (2) a. A lady is well behaved b. A lady must be well behaved The second aim of the paper consists of providing a new modal analysis for ISs in which we reconsider the relations between facts and their categorization. To account for this relation, we use two parameters: worlds and 'respects'. The framework that we propose allows us to account for the fact that, contrary to what theoreticians have claimed, ISs are not analytic statements. To state it briefly, the core argument in defense of the traditional view is the following. There are two distinct ways to obtain the generic reading of a sentence. The first is by directly making reference to a kind, as in (3), where the definite singular denotes the kind 'lion'. (3) The lion has a mane The second way is to use characterizing sentences, where the generic reading cannot be identified with a particular constituent, but rather belongs to the whole sentence. Characterizing sentences have been argued to be of Overt and covert modality in generic sentence