Je soutiens que Kripke n'a pas réussi à montrer que certains principes plausibles gouvernant l'attribution de croyances, tels que les principes de décitation et de traduction, pouvaient nous conduire à attribuer des croyances de dicto contradictoires à un sujet réfléchi et linguistiquement compétent sans présupposer une théorie descriptive des noms propres ou des termes désignant des espèces naturelles. Les cas décrits par Kripke se réduisent à des variantes du problème de Quine concernant les croyances de re ou à des variantes (...) du puzzle de Mates concernant l'intersubstiruabilité des synonymes dans les contextes de croyance.I claim that Kripke failed to show that certain plausible principles governing belief ascription, such as the disquotation and translation principles, could lead one to ascribe contradictory de dicto beliefs to a reflexive and linguistically competent agent without presupposing a descriptive theory of names or natural kind terms. Kripke's examples are nothing but variants either of Quine's problem about de re beliefs, or of Mates' puzzle concerning the interchangeability of synonyms in belief contexts. (shrink)

The phenomenon of quantification into attitude ascriptions has haunted broadly Fregean views, according to which co-referential proper names are not always substitutable salva veritate in attitude ascriptions. Opponents of Fregeanism argue that a belief ascription containing a proper name such as ‘Michael believes that Lindsay is charitable’ is equivalent to a quantified sentence such as ‘there is someone such that Michael believes that she is charitable, and that person is Lindsay’. They conclude that the semantic contribution of a name such (...) as ‘Lindsay’ is the same as the semantic contribution of a variable under an assignment, which these opponents suggest is merely the object assigned to that variable. However, renewed interest in variables suggests that they make a more complicated contribution to the semantic processing of sentences that contain them. In particular, a variable contributes both an assignment-unsaturated and an assignment-saturated semantic value. I use this dual role of the semantics of variables to develop a response to the argument from quantifying in. I take as my point of departure Cumming's () view that an attitude ascription relates the subject of an attitude to the assignment-unsaturated semantic value of an open sentence. I argue that this approach fails. I propose an alternative, according to which the truth of a belief ascription depends on both the assignment-saturated and the assignment-unsaturated semantic value of the open sentence in its that-clause. This approach reverses standard assumptions concerning the relation between quantification and substitution. (shrink)

‘Linguistic meaning must be public’ – for Quine, here is not a statement to rest with, whether it be reckoned true or reckoned false. It calls for explication. When we do, using Quine’s words to piece together what he thought, we find that much too much is concealed by the original statement. Yes, Quine said ‘Language is a social art’; yes, he accepts behaviourism so far as linguistic meaning is concerned; yes, he broadly agrees with Wittgenstein’s anti-privacy stricture. But precisely (...) what is being said by the original statement to be public, and what does calling it ‘public’ amount to? Pressing such questions complicates the picture enormously, partly though by no means entirely aligning Quine with linguistic internalism vis-à-vis Chomsky. (shrink)