UMR 7320 : Bases, Corpus, Langage

A Semantics of attitude reports based on a novel ontology of attitudinal objects

Friederike Moltmann
(directrice de recherche au CNRS, actuellement chercheuse invitée à l’Université Ca’Foscari de Venise)

Résumé

The standard view of the semantics of propositional attitude reports such as (i) centers on the notion of a proposition, an abstract object that serves as a bearer of truth conditions, as the meaning of sentences, as the shareable content of mental attitudes, and an argument of two-place attitudinal relations:

(i) John believes that Mary is happy.

The standard analysis thus takes (i) to have the logical form in (ii) with the that-clause standing for a proposition:

(ii) believe (John, [that Mary is happy])

Abstract propositions raise a range of problems, though, such as how they can be grasped and play a role in the mental life of agents if they are abstract objects, and how, as abstract objects, they can represent and be true or false. Furthermore, how come natural language hardly allows for explicit reference to propositions with its non-technical terms? I will outline an alternative view of attitude reports, which centers on the notion of an attitudinal object, things we refer to as ’beliefs, ’claims’, ’judgments’, ’requests’, ’desires’, ’intentions’, and ’decisions’. Attitudinal objects are extremely well-reflected in natural language, in particular in nominal constructions (the belief that S, the decision to do X), in complex attitudinal predicates (have a belief that S, make the decision to do X), and in the actual semantic behavior of quantifiers like something, which can take the place of clausal complements. Atttudinal objects thus are part of the ontology reflected in natural language, the ontology speakers implicitly accept when they use the language. Attitudinal objects moreover are part of a cognitively realistic ontology as mind-dependent entities that come with representational abilities, thereby avoiding the sorts of problems abstract propositions raise. I will argue that those two types of considerations are important when developing a semantics of attitude reports and I will outline a range of perspectives that attitudinal objects raise for research in syntax and semantics, in cognitive science, and in typology.