Back to list

Detail of contribution

Auteur: Aynat RUBINSTEIN

Co-Auteur(s): Paul PORTNER, Georgetown University, USA

Titre:
Contextual commitment and desiderative verbs


Abstract/Résumé: Languages express a variety of concepts of desire, e.g. the English trio hope, want, and wish. There has been much work on these items, mainly using a possible worlds analysis of attitudes. Following Heim 1992, a major theme is that desire verbs differ in their relation to the subject's beliefs: x wants p expresses x's preference for p with respect to x's beliefs, while wish expresses a preference for p with respect to a revision of x's beliefs to include possibilities in which p is true. Others argue that hope differs from want in that hope entails that the complement is compatible with the subject's beliefs, while want does not (e.g., Anand&Hacquard 2012). We build on this literature, add a concept of contextual commitment (see Rubinstein 2012, Portner&Rubinstein 2012), and argue that it explains additional distinctions in meaning among desire verbs: A is CONTEXTUALLY COMMITTED towards modal background h in w if A would argue for h(w) in a conversationally appropriate way in any relevant conversation C (by arguing that it is sensible/appropriate/wise). Following Portner&Rubinstein, hope requires the subject to be contextually committed towards its doxastic and buletic modal backgrounds, while the subject of want need not be committed to either one. This explains why one can want things one believes impossible, (1), or conversationally indefensible, (2): (1) I want/*hope to build a perpetual motion machine. (2) [Your doctor has told you that if you continue to smoke, you will soon die.] Gosh, Doctor, with all this stress I really want/?hope to have a cigarette now! For (1) to be true, the modal base must reflect a revision of the subject's beliefs, to which she is not committed (she would not argue that they represent the facts). With (2), the buletic ordering source is not defensible given what the doctor said. Contextual commitment also explains the characteristics of wish. Wish is very close to want when it takes an infinitival, but has its characteristic "counterfactual" meaning when its complement has "subjunctive" past marking: (3a) I want/?wish to smoke a cigarette later. (3b) I wish that John were here. We propose that wish requires commitment towards its buletic ordering source; this is why wish (like hope) is unacceptable in the doctor scenario. In (3b), wish revises the doxastic modal base to make it compatible with John's being here, and the subject's desires are defensible with respect to the expanded belief set. Finally, hope differs from wish in requiring commitment to the modal base (disallowing belief-set revision). In the talk we will also discuss connections between these ideas and counterfactual marking in 'wish'-constructions (Iatridou 2000, von Fintel&Iatridou 2008) and performative uses of desire verbs.