Abstract

Davidson's paratactic account of indirect speech exploits the fact that ‘that’ can be either a demonstrative pronoun or a subordinating conjunction. Davidson thinks that the fact that it is plausible to think that it inherited the latter function from the former lends support to his account. However, in other languages the two functions are performed by unrelated words, which makes the account impossible to apply to them. I argue that this shows that, rather than revealing the underlying form of indirect reports, the account reflects only a quirk of English.