Integrated parentheticals in quotations and free indirect discourse

Free Indirect Discourse and Quotations are introduced by a special kind of parentheticals, minimally constituted by a subject and a predicate of saying, thinking etc. I propose that these parentheticals are represented in a syntactic structure integrated with that of the reported sentence. The phrases hosting the parentheticals are projected by prosody oriented heads, i.e., heads which do not express a lexical content, but are read at the syntax-phonology interface as special instructions for realizing the peculiar intonation associated to parentheticals, the so-called comma intonation. I show that this approach offers several advantages, contributing in solving some long standing problems connected with the syntactic status of parentheticals in general.

In this chapter I consider the syntactic properties of a particular kind of parentheticals, those introducing Quotations – henceforth, QU – and Free Indirect Discourse – henceforth, FID. Consider the following examples: (1) I will leave tomorrow, said John (2) The new ration did not start till tomorrow and he had only four cigarettes left, thought Winston (adapted, from Orwell 1984). Example (1) is a QU structure and the parenthetical in question is said John. Example (2) is a FID construction and the parenthetical is thought Winston. As already well known, they have special properties from an interpretive, syntactic and phonological point of view. QU and FID parentheticals are alike under many points of view, even if the two constructions must be kept separate, especially with respect to the interpretation of pronouns and verbal forms. For the purposes of this work, I will in general consider them alike. Observe now the following paradigm: (3) John said that Mary left (4) John said: “Mary left” (5) Maria, said John, left. It seems to me that the most important goal for a syntactic analysis is to provide a coherent analysis of the similarities and differences among the constructions in (3)- (5). At first sight, these structures seem very much alike, both from the point of view of their meaning and their syntax – to the extent that some scholars have proposed a direct syntactic derivation (Emonds 1973; Ross 1973), for instance of (5) starting from (3). I will show here that the situation is indeed much more complex than that. In particular, in this paper I show that example (5) is closer to (4) than to (3). The approach I will develop here is an integrated view of parentheticals, complying with Kayne’s (1994) Linear Correspondence Axiom (LCA).