On saying that again

One of Donald Davidson's important contributions in the philosophy of language is his paratactic account of the semantics of indirect speech reports, presented in his classic 'On Saying That' [2]. In their recent paper 'You Can Say That Again' [6], Ernest LePore and Barry Loewer have attempted to generalize the paratactic account so as to serve in the analysis not only of indirect speech reports, but also of such constructions as believed that, desired that, and by implication all members of an intuitively reasonably well-defined class of constructions including expected that, doubted that, denied that, and such related cases as wanted NP to and said for NP to, including cases where the NP is null, as in wanted to and said to. I shall argue that there is a rather extensive variety of data from English that the paratactic account, as generalized by LePore and Loewer, is incapable of accommodating; indeed, some of these data involve said that, and so count against Davidson's own more limited proposal. In Section 1 I exposit Davidson's paratactic account of indirect speech, and in Section 2 1 look at some objections to it. LePore and Loewer's extension of the account is described in Section 3. Their extension is subject to various further objections which do not apply to Davidson's narrower proposal. These are discussed in Section 4.