| This thesis investigates the syntax, semantics, and pragmatics of cleft sentences in English. Following an introduction to critical discourse-pragmatic notions (e.g., 'topic-comment,' 'cognitive status,' and 'accent'), the first part focuses on the syntactic and semantic nature of the four cleft subparts shown below and the relation between them:(UNFORMATTED TABLE OR EQUATION FOLLOWS);The second part focuses on the discourse-pragmatics of clefts, drawing on a corpus of 700 spoken and written tokens. Two discourse-pragmatic subtypes are distinguished: 'topic-clause' clefts with primary accent on the clefted constituent, and 'comment-clause' clefts with primary accent on the cleft clause. Gundel's (1985) generalization that sentence-final topics are necessarily activated is supported by the consistently activated status of topic cleft-clauses, and is extended by a classification into subtypes of direct and indirect activation. It is argued that Gundel's label 'comment' is preferable to Prince's (1978) label 'informative' as a characterization of primary-accented cleft clauses, and that the clefted constituent sometimes expresses the sentence topic. Generalizing the view that topic-clause clefts answer contextually-relevant questions, it is suggested that comment-clause clefts answer multiple questions.;It is argued that the cleft clause is simply a relative clause which is syntactically linked to the cleft pronoun, an adaptation of Akmajian's (1970) 'extraposition' analysis of clefts. Cleft pronouns are shown to vary in form in colloquial 'th-clefts' (Ball 1978), depending on the cognitive status of the open proposition denoted by the cleft clause, and to agree in number with the clefted constituent in predicational but not in specificational clefts, supporting the view of Bolinger 1972 and Gundel 1977 that the cleft 'expletive' has denotational content. |
