Chapter One Introduction
1.1 Background of Research
The nature of language is change. Language change is a universal and continuousprocess，evidenced by such phenomena as the changes between British English andAmerican English, and the evolution of ancient English to modern English.Grammatical change is pervasive and systematic，and it may follow different patternsin different sub-registers of the language.Over the past few decades, there has been increasing interest within the field oflinguistics in the study of how languages change. Many aspects of grammaticalchange have been studied，such as the changing use of the subjunctive mood，modalauxiliaries，semi-modals，progressives，the passive voice, non-finite clauses，and so on.Several models for explain changes in language have been developed, such as those ofgrammaticalization，colloquialization, densification and Americanization (Leech,2009).Barber (1964) conducted an early study of changes in the English language. Hefocused mainly on changes in pronunciation, the growth of the vocabulary, changes inmeaning and grammatical changes. Foster (1968) focused his research on thegrammatical differences between British English (BrE) and American English (AmE).He argued, for example, that BrE was not following AmE in replacing the presentperfect (PP) with the simple past (SP) in the 1960s. Christian Mair and GeoffreyLeech (2006) fiirther researched the changes occurring in the English language，with afocus on recent changes in English syntax.
………..
1.2 Objectives of Research
The research attempts to answer the following questions;
1. Are there any changes in perfect aspect in American English in recenttwenty-two years?
2. If a change in perfect aspect does exist, is the frequency of using perfectconstruction increased or decreased during the past two decades?
3. What is the pattern or trend of this change? Is it changed smoothly or withouta stable pattern?
4. If a change in perfect construction does exist, which category contributed tothe change in perfect construction? Is it past perfect, present perfect orinfinitival perfect?
5. If a change in perfect construction does exist，which register contributed tothe change in perfect construction? Is it spoken, fiction, magazine, newspaperor academic?
6. What caused the changes in each register? Is it colloquialization orgraiTimaticalization?
………..
Chapter Two Literature Review
2.1 Aspect
The term aspect is less familiar to the public than are terms such as tense or mood.There are several aspectual subsystems such as viewpoint aspect, situation aspect andphasal aspect. It is impossible to describe one aspectual subsystem without makingreference to another. Situation aspect, (also called Aktionsart or lexical aspect), refers to a semanticproperty of predicates which depends on the meaning of the verb and properties of itsinternal structure. Situations can be divided into two categories: with an inherentendpoint or without. For example, "He is a good guy" is a state, while "He had dinnerin home this morning" is an event with an endpoint.Vendler (1967:102-121), made a classification of four types of situation type:accomplishments, achievements, activities, and states. However, some linguistsbelieve that there is a fifth type of situation. Smith (1997) called this new typesemelfactives. He concluded that the first type, which called states, has the property of Static and durative. The examples of this type are “like this girl", or “know him,，. Thesecond type, which called activities, has the property of dynamic, durative and atelicevents. The examples of this type are like “walk in the street", “carry a box”. Thethird type is called accomplishments, and it has a property of dynamic, durative, telicevents consisting of a non-detachable process with an outcome. The examples of thistype are like “run to school”，“read a book" and “build a bridge". The fourth type isachievements, and its property should be dynamic, telic and instantaneous events. Theexamples of this type are “fall in love", “reach the city". The new type semelfactiveshas the properties of dynamic, atelic, and instantaneous events. The examples of thistype are “tap"knock" and "flap a wing".
………
2.2 Perfect
As discussed in chapter 2.1, perfect，as opposed to progressive, is an aspectualcategory. Combined with tenses, we can get three categories: present perfect (PrP),past perfect (PaP) and future perfect (FuP).Based on Reichenbach's (1947) system of temporal relations，Michaelis (1998:107) concluded the tense and aspect component of these three categories ofsentences. R means reference time, the temporal standpoint or perspective from whicha situation is presented, S stands for speech time, the moment of speech, and E meansevent time, the time at which an event or state occurs or holds. The relationshipbetween these three terms is illustrated in the follow table. The symbol and 'means the relation of a temporal precedence. The first sentence is present perfect, the speech time and reference time are all“now”，but the event time is before the reference time, so it is in perfect construction.The second sentence is past perfect, the reference time (came) is before the speechtime (now)，and the event time (left) is before the reference time. The third sentence isfuture perfect, the reference time is after the speech time，and the event time is beforethe reference time.
……….
Chapter Three Research Methodology and Data Collection ........ 21
3.1 Research Material: COCA....... 21
3.2 Research Instruments .......23
3.2.1 Research Instruments: Excel....... 23
3.2.2 Research Instruments: Stata....... 24
3.3 Research Procedures and Processing....... 26
Chapter Four Results and Discussion........ 32
4.1 Diachronic Analysis of the Changes in the Past Two Decades....... 32
4.2 Chi-squared Goodness of Fit Test: Is there Change? ....... 35
4.3 Time Series Analysis ....... 41
4.4 Changing Trend in Different Registers .......47
4.5 Interpretation of the Changing Trend in Different Registers....... 60
Chapter Five Conclusion ....... 62
5.1 Summary of Research Findings....... 62
5.2 Limitation of the Study .......64
Chapter Four Results and Discussion
Data on frequencies of perfect in the COCA corpus over the past twenty-twoyears, collected through procedures described in Chapter Three, enabled the analysisof the change pattern in the use of perfect in modem English. Chapter Four reportsvarious results coming out of this analysis. Section 4.1 provides summary of thecollected data sample, and offers diachronic analysis of the changes over time from1990 to 2012. Section 4.2 uses various Chi-squared tests for goodness-of-fit toexamine whether the pattern of use of perfect in later time periods is different to thepatterns in earlier time periods. Section 4.3 uses structural break model and linearregression model to look at the time series pattern more rigorously, in order toexamine whether there is a continuous downward trend or whether there exists a“structural break", i.e. whether the use of perfect in English experiences a sharpchange in some years and then settles into another regime in other years. Section 4.4explores the different changing pattern in five registers: spoken, fiction, magazine,newspaper, and academic. By looking separately at different registers of English, thethesis can identify whether the change in use of perfect is uniform across differentdomains in English language, and can seek out the contributions of different domainsin English in the change in use of perfect.
……….
Conclusion
A significant decreasing trend in usage of the perfect construction in AmericanEnglish during the past twenty-two and a half years was found in this thesis. Thefrequency of use dropped rapidly from 1990 to 1997，maintained relative stabilityfrom 1998 to 2007, and then dropped again from 2008 onwards. The PMW ratiodropped 12.8%, from 4359 in 1990, to 3801 in 2012. In terms of the four categories ofperfect construction，two showed continuous rates of decrease in use: the presentperfect and the nonfmite perfect. Usage of the present perfect dropped by about 20%in recent twenty-two and a half years，and usage of the nonfmite perfect dropped by29%. Chi-squared testing confirms that the observed changes are statisticallysignificant. No pattern of change was observed for usage of the past perfect andinfinitival perfect. Linear regression analysis showed that usage of the past perfectincreased over time, which balanced out some of overall decrease in overall usage ofperfect constructions, while usage of the infinitival perfect dropped slightly over thepast two decades. A smoothing technique was applied in order to reduce noise in orderto get a clearer picture of the changing trend. To summarize, a decelerated decreasewas found in usage of perfect constructions, a trend of accelerated increase was foundin usage of both the nonfinite and present perfect, usage of the infinitival perfect wasfound to be quite stable, and usage of the past perfect evidenced a ‘‘LT shape curve.
…………
Reference (omitted)
