Analysis Of Research Paper Abstracts From Functional Perspectives

Discourse analysis has been developing quickly since 1960s. The scholars at home and abroad have made a great contribution to this field, much functional analysis of various texts or genres have been carried out in academic and professional settings. The particular genre of research paper (RP) has received much attention. However, there is limited genre analysis on research paper abstracts (RPA). That is to say, such study is not so ample and thorough. So this thesis attempts to study respectively the four aspects of RPA (purpose, method, result and conclusion) in the view of ideational function, interpersonal function and textual function. It tries to explore the characteristics of RPA from the angle of the cohesion of lexicon, the function of grammar, the realization of semantics and the distribution of information. Hopefully, this thesis could help researchers learn about more information within least time and give them guide in writing papers.The main contents of this thesis extend from chapter 3 to chapter 5. These chapters give more detailed explanation and analysis of transitivity, mood and modality, tense and voice, theme and thematic progression, and grammatical metaphor (nominalization). The following is the results of the research:(1) In our corpus, material process is predominant, the ratio reaches 56%, and they can be seen easily in the stage of Method and Result of an abstract. Relational process is just after it. It distributes mainly in the stage of Result with 25% frequency. The following processes are verbal process (12%) and mental process (5%), all mainly happen in the stage of Result.(2) Almost no mood adjuncts are found in the chosen corpus except for"only", which appears once in text 22. Quite adverbs working as mood adjuncts exist in the data, such as significantly, successfully, randomly, regularly and so on.(3) Most modality adjuncts can be found in the corpus, but the frequency of reusage is different. Could (46%) is much higher than others, the following is can (23%), may (15%), might (8%), will (5%), must/should (5%, respectively).(4) In the corpus, the past tense is key one with 63% ratio, but its distribution in the four stages of RPA is uneven, we could find mainly in the stage of Method and Result. The next is present tense, which runs through all stages of abstract with 35% ratio.(5) By comparison of active voice with passive voice, active voice in RPA is still...