Mid-Term Report Summary - BLENDS (Between Direct and Indirect Discourse: Shifting Perspective in Blended Discourse)
The received view in semantics, also familiar from prescriptive grammar textbooks, is that there are two, fundamentally distinct modes of reported speech: direct discourse (Mary said, “I'm crazy”) which involves a form of quotation, and indirect discourse (Mary said that she's crazy). The aim of the BLENDS project is to challenge this view. In the first half of the project I proposed a mechanism called mixed quotation to unify both direct and indirect discourse. In this half I refined this approach, adding an analysis of unquotation and demonstration. This approach allows us to describe a host of phenomena that would fall outside the classical dichotomy.
BLENDS has investigated this gray area between direct and indirect discourse also empirically. Specifically, in this second half we looked for hybrids in (i) ancient Greek (where we found that authors shift freely from direct to indirect and back); (ii) modern literary narratives (where we analyzed Free Indirect Discourse as a kind of direct discourse); (iii) sign language (where investigated how native signers use Role Shift to represent others' signs, words, or actions); and (iv) child language (where we found that Dutch children up to age 10 have difficulty processing direct speech, despite the fact that they spontaneously produce direct speech well before indirect speech).
https://sites.google.com/site/emarmaier/blends/