Pragmatics, subjectivity and the grammaticalization of the English perfect

The dissertation provides a gradual account of the development of a 'have' perfect from a secondary predicate/adjectival construction. The thesis is founded on the theoretical notion that semantic/change is brought about by the conventionalization of conversational implicatures (cf. Traugott and Konig 1991). The analysis is based on data taken from four periods in the history of English: Alfred (c. 850), AElfric (c. 1050), early Middle English (c. 1225), middle Middle English (c. 1375). Frequency data regarding the semantics of the verb participle and accompanying adverbs are calculated to help identify particular semantic-pragmatic contexts that bring about a semantic shift from one stage to the next.; Chapter 2 focuses on the early development of the perfect from a secondary predicate to a resultative construction, claiming that the early Old English {dollar}have + participle{dollar} construction was still a type of resultative. Chapter 3 accounts for the shift from resultative {dollar}to{dollar} perfect, arguing that the shift was facilitated by perception verbs and iterative/durative contexts. The analysis presented for the historical English data is further supported by dialect variation in the use of the tener + past participle construction in Modern Castilian Spanish.; Chapter 4 relates the diachronic analysis presented in Chapters 2 and 3 to the four use types typically posited for the Modern English perfect, resultative, experiential, continuative and "hot news". It illustrates the diachronic paths of extension between the different use types and argues that the non-discreteness of the use type categories can be traced to the diachronic/pragmatic process from which they arose. Using data from the Los Angeles Times and CBS Special Reports, the chapter also argues that the Modern English perfect is pragmatically linked to two other uses cross-linguistically attested within the category perfect: recency and modality. Chapter 5 examines two different notions of subjectification in grammaticalization (cf. Traugott 1989 and Langacker 1990) by applying them to grammaticalization paths attested within the category 'perfect', in particular, the shift from resultative to perfect.