par Bochner, Gregory 
Référence Lecture notes in computer science, 10257 LNAI, page (3-14), (forthcoming)
Publication Publié, 2017
Référence Lecture notes in computer science, 10257 LNAI, page (3-14), (forthcoming)
Publication Publié, 2017
Article révisé par les pairs
| Résumé : | In this paper I sketch an alternative to Stalnaker’s view of referential uses of descriptions. Stalnaker has long promoted a pragmatic account of assertions, presuppositions, and informativeness. He is also a fervent advocate of propositionalism, the doctrine that the contents of assertions, presuppositions, and attitudes, (are or) determine sets of possible worlds. I argue that the combination of a pragmatic account and propositionalism creates several problems. (i) It does not predict the right truth-conditions for some assertions. (ii) It cannot duly separate facts of reference from presuppositions about facts of reference. (iii) It reproduces, at the level of what is presupposed, the cognitive significance problems that pragmatic presuppositions were meant to solve at the level of what is asserted. I argue that the solution to these problems involves giving up propositionalism. While Stalnaker analyses assertions and presuppositions in terms of singular propositions and possible worlds, I propose to analyse them in terms of properties and centred worlds. But unlike other centred world accounts inspired by Lewis, the view I advertise is not egocentric: the circumstance of evaluation of an assertion need not be centred on the subject, it can be centred on an object. When the assertion involves a referential use of a description, the object at the centre is the one that the speaker “has in mind.” Unlike its egocentric counterparts, this view can maintain that referential communication is direct: speakers and hearers can grasp the same truth-conditions. |
