Propositional Attitudes and Opacity

In belief and other attitude reports, swapping co-referring names can change truth value, creating opaque contexts that strain the classical logic of identity.

Definition

A context is referentially opaque when substituting co-referring expressions within it can change the truth value of the whole; propositional-attitude verbs typically create such contexts, posing a problem for the classical principle of substitutivity of identicals.

Scope

This topic covers the logic and semantics of propositional-attitude reports such as 'believes that', 'knows that', and 'hopes that', and the referential opacity they exhibit. It treats the failure of substitution of co-referring terms and of existential generalization in attitude contexts, Frege's resolution by reference-shift to senses, Quine's distinction between notional and relational (de dicto and de re) attitude ascriptions and his worries about quantifying in, and Kripke's puzzle about belief.

Core questions

• Why does substitution of co-referring terms fail in belief reports?
• How should attitude verbs be analyzed semantically?
• Can we quantify into attitude contexts, and what would de re belief require?
• Does the opacity reflect logical form, sense, or the structure of mental states?

Key concepts

• referential opacity
• substitutivity of identicals
• de dicto vs. de re attitudes
• quantifying in
• customary vs. indirect reference
• Kripke's puzzle about belief

Key theories

Fregean reference-shift

Frege handles opacity by holding that within an attitude report an expression refers not to its ordinary reference but to its customary sense, so co-referring terms with different senses are not genuinely intersubstitutable there.

Notional vs. relational ascriptions

Quine distinguishes de dicto (notional) from de re (relational) attitude reports and argues that only the latter permit quantifying in, while raising doubts about the intelligibility of de re belief.

History

Frege's 1892 treatment of indirect reference set the agenda. Quine's 1956 paper sharpened the de dicto/de re distinction and the problem of quantifying in, and Kripke's 1979 puzzle about Pierre's beliefs concerning London showed that the difficulties arise even without substitution, challenging both Fregean and direct-reference accounts.

Debates

How to reconcile direct reference with opacity

Whether attitude reports require Fregean senses to explain substitution failures, or whether a direct-reference (Millian) semantics can accommodate the data by appeal to guises, modes of presentation, or pragmatic factors, as Kripke's puzzle presses.

Key figures

• Gottlob Frege
• W. V. O. Quine
• Saul Kripke
• Nathan Salmon
• David Kaplan

Seminal works

• frege1892
• quine1956
• kripke1979puzzle

Frequently asked questions

What is referential opacity?

A linguistic context is referentially opaque when you cannot freely substitute terms that refer to the same thing while preserving truth. 'Lois believes Superman can fly' may be true while 'Lois believes Clark Kent can fly' is false, even though Superman is Clark Kent, because the belief context is opaque.