|
The main purpose of this paper is to investigate the relative information status of the preposed and postposed constituents of inversion in English, based on the notion of discourse familiarity presented by Birner (1994). The preposed constituent and the postposed constituent are abbreviated as PRE and POST respectively in this paper. The study shows that there can be three types of combinations of PRE and POST in discourse familiarity: old-new, old-old, and new-new, while there is not a single token in which PRE is discourse-new and POST is discourse-old. This result confirms the argument given by Birner (1994) that PRE in a felicitous inversion must not represent less familiar information in the discourse than does POST.
In order to explain the cases in particular where the two constituents have the same discourse familiarity, I will support a gradient notion of information status rather than its given/new binary distinction and argue that the information of POST is more important than that of PRE. As the relevant factors for determining relative discourse familiarity or importance of information, I will adopt recency of mention and topichood.
|