A Corpus-based Analysis Of Recent Changes In American English Perfect Construction

Language changes all the time. The emergence of various corpora, such as the Brown family corpora, has made it possible for researchers to study the changes in the English language in a quantitative way. Studying language change from the perspective of aspect change is always fruitful, but most of the studies in this area focused only on British English, or spoken register. This thesis tries to fill some blanks in this area.This thesis is aimed to find out the recent changes of perfect aspect in American English from the register perspective. The research material used in this paper is COCA (The Corpus of Contemporary American English), the largest and balanced corpus of American English. The raw materials of over450million words from1990to2012were processed. The numbers of tokens in all four categories of perfect (present perfect, past perfect, nonfinite perfect and infinitival perfect) and in each of the five registers (Spoken, Fiction, Magazine, Newspaper, and Academic) were calculated in Excel and processed through linear regression model and Chi-square test in software Stata.A decreasing trend in using perfect aspect is found through data analysis. The main contributor to this decreasing trend is present perfect construction. Because of colloquialization and drift to spoken English, among all five registers, Spoken, Magazine and Academic register experienced a significant decrease in the use of the perfect construction. There is no significant change in Newspaper register because of the joint effect of colloquialization and grammaticalization, while there is an increase in the use of perfect construction in Fiction register because most of the register was past perfect.This thesis proved that English aspect change can be learned from the register perspective and our results may throw some light into English teaching by showing the recent English aspect changing patterns.