Chapter 1 Introduction

1.1 Research Background

Nowadays English has increasingly become an international language，especiallyin science and technology. This can be proved by the fact that in the scientific world，most academic writings are in English and a large number of world-known journalsonly receive research articles written in English. Therefore，English academicdiscourse has become an important means for researchers and scholars worldwide topresent their research achievements, put forward and publicize their ideas, andcommunicate with others in the academic community.As the number of English academic writings published in international journals，conferences and symposiums keeps increasing，more and more scholars have realizedthat English used in the academic world is different from that used in daily life or inbusiness and trade. Being a branch of English for Specific Purposes，English used inthe academic world (or English for Academic Purposes) have gradually attractedattention from scholars from about two decades ago. Since then，academic discoursein the written form like research articles，master theses, PhD dissertations, conferencepapers, research reports, memos，etc. has been one of the research focuses ofdiscourse analysis，and influencing studies include Bazerman (1988)，Swales (1990,2004), Halliday (1987，1990) and others.

……….

1.2 Research Questions

The analyzing model that will be adopted in this thesis is grounded in twotheories in academic discourse studies. One refers to genre analysis, which can beapplied to uncover the macro structure of the Conclusion section. The other is thesystemic functional grammar, which will be used to analyze the micro linguisticfeatures of the Conclusion section. By combining both perspectives, this study is designed to answer questions as follows:

(1) What is the macro structure of the Conclusion section? Does it conforms toYang & Allison's (2003) model or not? And why?

(2) What are the micro linguistic features in the Conclusion section?

(3) In terms of the transitivity system, which processes are the typical ones in theConclusion section and what are these processes trying to express?

(4) In terms of the modality system，what are the commonly used modaloperators in the Conclusion section and what are their functions?

(5) In terms of the Theme system，which kind of Theme is frequently used? Whatare the internal structures of these Themes?

……….

Chapter 2 Literature Review

2.1 Studies of Academic Discourse Overseas

Many studies on academic discourse since the 1980s have been undertaken froma variety of perspectives (Biber et al.，2002). According to Swales (2002: 150-164)，approaches to academic discourse analysis are traditionally regarded as "top-down",that is，they are mostly concerned with whole，individual texts as well as the socialconditions of their production and reception. Charles et al. (2009: 1) identifies at leasttwo themes in traditional discoursed-based analysis of academic writings and pointsout that these approaches to academic discourse analysis can be generallydifferentiated by the ways they interpret and balance these themes.One theme concerns academic writings in terms of 'moves', the recurringpatterns of meaning in texts of similar types (Charles et al., 2009: 2). Studies of thiskind adopts a rhetorical or social/historical perspective towards academic discourseanalysis, which investigates not only the rhetorical structure of academic discoursebut also the way the practices of researchers in particular discourse communitiesshape the conventions of academic genres (Biber et al., 2002: 11). Swales' (1990)pioneering CARS (Create-A-Research-Space) model of research article (RA)introductions has generated numerous investigations of this kind in the field of genreanalysis. Based on that，various researches have been done concerning all parts of RAsections (Nwogu, 1989; Samraj，2002，2005; Brett，1994; Williams, 1999; Hopkins &Dudley-Evans, 1988; Holmes, 1997).

……….

2.2 Domestic Studies of Academic Discourse

Researches on academic discourse in China are about two decades later thanthose in the West. Unlike the tripartite picture of academic discourse analysis overseas，domestic studies of academic discourse mainly centered on the "top-down"approaches，genre analysis pioneered by Swales (1990) and SFL-based analysisdeveloped by Halliday (1994)，Hasan (1984)，Martin (1992) and others. The"bottom-up" methodological approach, MD analysis initiated by Biber (1988)，is formost part rare. Only two research articles published in CSCI (China Science CitationIndex) journals are found to introduce Biber's MD analysis or apply it to academicdiscourse (Lv，2010; Lv & Zhou，2013). Researches based on Swales' (1990) genre analysis are largely empirical studiesthat apply move-step analysis to the discussion of academic discourse and fewcontribute to the theoretical basis of genre analysis. One notable feature of thesedomestic studies is that nearly all of them focus on only one part-genre of RAs ormaster theses “"“the Abstract section, either attempting to examine disciplinarysimilarities and differences in these abstracts, or trying to draw a contrastive analysisbetween English and Chinese abstracts. Ge & Yang (葛冬梅、杨瑞英，2005) identifiesa five-move structure in their corpus of abstracts on the basis of Bhatia's (1993)four-move model and discusses three linguistic features (model verbs，tense andfirst-person pronoun) in relation to disciplinary characteristics. On the contrary, Ye &Liuattempts to figure out the common features in abstractsacross different disciplines and argues that teaching the common core in academicdiscourse across disciplines helps students to explore similarities and variations inlanguage use across discipline from critical and comparative perspectives. Kang &Sun (康勤、孙萍，2012) also examines both the similarities and differences inabstracts across disciplines.

………..

Chapter 3 Theoretical Framework ......  20

3.1 The Conclusion Section: A Neglected Research Area......  20

3.2 A Tentative Macro-and-Micro Analyzing Model ...... 21

3.2.1 Genre Analysis and Yang & Allison's (2003) Model......  22

3.2.2 Three Metafunctions of Language ......   26

3.3 Summary ...... 34

Chapter 4 The Macro Analysis of RA Conclusion Sections......  35

4.1 Results of the Macro Analysis of RA Conclusion Sections......  35

4.2 Move 1 in the Conclusion Section......          36

4.3 Move 2 in the Conclusion Section......  39

4.4 Move 3 in the Conclusion Section......  40

4.5 Structural Differences across the three Journals......  42

4.6 A Brief Summary of the Move-Step Analysis......        43

Chapter 5 The Micro Analysis of RA Conclusion Sections ...... 45

5.1 The Transitivity System in the Conclusion Section......  45

5.2 Modality in the Conclusion Section ...... 51

5.3 Themes in the Conclusion Section......        55

5.4 Summary ......           62

Chapter 5 The Micro Analysis of RA Conclusion Sections

5.1 The Transitivity System in the Conclusion Section

The transitivity system，which is considered as the most important system withinthe category of the experiential function, represents human experience in clauses bymeans of processes. In other words，reality is made up of different process types(Halliday, 1994: 106). Altogether，there are six process types，i. e, material，mental，relational，behavourial, verbal and existential. Based on this，the 40 Conclusionsections are divided into the level of clauses which are then classified into differentprocess types. The following table shows an overview of the results.

…….

Conclusion

Academic discourse analysis in the last two decades has mainly been done fromthree perspectives: a genre analysis which uses a move-step analysis; MD analysiswith the help of corpus; and SFL-based discourse analysis. Studies adopting theseperspectives have produced fruitful research findings and far-reaching influence onthe teaching of English for Specific Purposes. Regretfully, previous researches havetraditionally focused on IMRD (Introduction，Method, Results and Discussion)sections of research articles or master and PhD theses，leaving aside closing sectionsof research articles like Conclusions most of the time neglected. This thesis presentsan investigation of the Conclusion section in 40 research articles.The analysis of the Conclusion section is based on a tentative macro-and-microanalyzing model that is built on genre analysis and SFL-based analysis. Examinedfrom the macro level，research findings indicate that the Conclusion section of RAgenerally follows a three-move structure first proposed by Yang & Allison (2003)，i. e.summarizing the study, evaluating the study and deductions from the research.

…………

Reference (omitted)