Savas L. Tsohatzidis

This paper intends to contribute to the evaluation of the project of analyzing speech act concepts in terms of mental state concepts, by examining Searle's and Vanderveken's proposed analyses of certain types of illocutionary acts as expressions of corresponding types of emotional states. It is argued that the proposed analyses are all defective, that the assumptions about underlying speech act/mental state parallelisms from which their initial plausibility might be taken to derive are themselves mistaken, and that the fact that they have been proposed at all betrays insufficient attention to the distinction between semanti-cally and pragmatically determined aspects of utterance interpretation. It is concluded that these results disturb both the overall structure of Searle's and Vanderveken's theory of speech acts, and the logic of several other similarly reductionist claims in the philosophy of language.

The most distinctive, and probably the most striking, assumption of Donald Davidson's well known ‘paratactic’ analysis of the logical form of saying ascriptions is that the “that”-clause that, in such an ascription, specifies the content of the ascribed act of saying, is neither syntactically nor semantically part of the sentence effecting the ascription. The present paper identifies a neglected problem that this assumption engenders for the Davidsonian analysis. The problem arises in connection with instances of saying ascriptions that are both self-verifying and self-referential, and consists in the fact that, given the Davidsonian assumption, these ascriptions must be assigned logical forms that misrepresent their logically relevant properties.