Indirect discourse

Abstract

This essay consists of a critical discussion of the main theories of indirect discourse. It is first argued that the theory put forward by Frege is fundamentally inadequate. Frege views expressions in indirect speech as standing for intensional entities. But application of the substitutivity rule in accordance with Frege's theory sometimes fails to preserve truth-value. There is the problem of quantifying simultaneously into a normal and an oblique context. It is shown that Frege's theory lends a spurious precision to the question of the conditions under which indirect quotation can be said to be successful. A truth theory meeting Tarski's criterion of adequacy is given for a simple Fregean language. It is argued that complex Fregean languages, i.e. those which match the expressive power of natural language, are not truth-theoretically tractable. The conclusion is drawn that an adequate theory must abjurereference to intensional entities and meet the demands of truth theory. Quine's theory meets the first of these requirements: he counts expressions in indirect speech as without semantic significance. But it is shown that Quine dissolves the semantic structure needed by a theory of truth. Both of the requirements are met by the theories of Geach and Davidson. Geach argues that oratio obliqua is logically superfluous and can everywhere be replaced with oratio recta. The arguments advanced against quotational theories prove to be fallacious. But Davidson's theory is prima facie more attractive than Geach's; for it counts oratio obliqua sentences as overtly exhibiting their logical form. Expressions in indirect speech play a normal semantical role: but are semanticallyinsulated from that to which we ascribe a truth-value. But it is shown that Davidson's theory as given is unable to deal with oratio obliqua sentences on their relational reading. In the last resort, Geach's theory seems the more viable of the two.