Cleft constructions in discourse
This thesis presents an analysis of the structure and function of cleft constructions in discourse. Drawing on a corpus of naturally-occurring spoken and written data, we present a multi-layered explanation of how it-clefts, wh-clefts, and reverse wh-clefts are different from non-clefts, and from one another. After a review of previous research on clefts in discourse, we explore the aspects of syntax, semantics, and pragmatics relevant to the structure and function of all three types of cleft. The discussion falls into three main parts: An analysis of the three cleft types, within the framework of Generalized Phrase Structure Grammar (cf. Gazdar et al. ), in which particular attention is paid to the variety of constituents that can appear in particular positions in each type. The output of the grammar rules is compared to the examples that occur in the corpus of data. A treatment of cleft presupposition in terms of an analogy (suggested by van der Sandt ) between presupposition and the treatment of pronominal anaphora in Kamp's  Discourse Representation Theory An examination of the range of accentual patterns, presuppositional relations, and information structures typically appearing in clefts of all three kinds. We show that marked distinctions exist between the three cleft types in terms of all these factors, and suggest ways in which this helps to differentiate the range of discourse contexts in which clefts in general, and each cleft type in particular, are appropriate. At the end of the thesis we point to an analogy between the formal model for clefts presented and a psychological model of sentence processing We also suggest how the conclusions regarding both the structure and function of clefts as a class of construction and the distinction between the three types of cleft could be synthesised in a decision procedure for syntactic choice. Finally, we suggest some related areas for further research.