A semantic theory T for a language L should assign content to utterances of sentences of L. One common assumption is that T will assign p to some S of L just in case in uttering S a speaker A says that p. We will argue that this assumption is mistaken.

Donald davidson has proposed an account of indirect discourse that has been the subject of a great deal of discussion. Critics have contended that the theory saddles sentences in indirect discourse with implications they do not have, That the theory rests on an unsuitably obscure primitive notion that it cannot be extended to "de re" constructions and that it cannot be extended to sentences about other propositional attitudes such as belief. In this paper, I formulate davidson's theory more precisely than (...) it has been formulated, Thus far, And defend the theory from all the criticisms just described. (shrink)

One commonly stated reason for thinking that there are abstract entities such as propositions is that they are needed to account for undeniable facts about propositional attitudes and the sentences reporting such attitudes. According to the propositional theory, belief, doubt, assertion and other attitudes are relations between individuals and propositions. In sentences reporting these relations the words in the content-sentence refer to concepts or other abstract things and the entire ‘that-clause’ refers to a proposition. According to a common rival account, (...) propositional attitudes are complex relations between individuals and sentences and each that-clause refers to the content-sentence it contains. The words in the content-sentence either fail to refer or refer to themselves. A striking implication of both the propositional theory and the sentential theory is that the words and phrases in the content-sentence of a sentence expressing a propositional attitude fail to refer to the familiar things to which they ordinarily refer. (shrink)