Macro-Discourse Analysis--a Scale-Orientated Model

Although as early as 1920s, the study of language in use has become the research interest of the Prague School and the London School, discourse analysis marks its birth as a modern discipline no earlier than the 1960s. (It Hi, 2002: 237; 1995: 14; van Dijk, 1997: 25) Up to now, it has witnessed two developmental periods, and in the second period it has taken these characteristics of a vast and multidisciplinary enterprise, involving at least half a dozen disciplines, like sociolinguistics, pragmatics, text linguistics, psychology, artificial intelligence, and anthropology, (2000: PI49; van Dijk, 1985: 1-9; latsko, 58) As a multi-discipline, its research can take several angles, which van Dijk (1997: 32) summarizes as the following three dimensions: (a) language use, (b) the communication of beliefs (cognition), and (c) interaction in social situations. Every text, in researches from those angles, is representative, and has the same effectiveness in terms of theoretical exemplification. But discourses are different and unique, if individuals' social and cognitive development is taken into account. Taking a functional view of language, this thesis is intended to exhibit discourse's uniqueness in relation to its context, that is, the activity, and to arrange discourses along two axes: one is the cognitive axis, based on pedagogical findings; the other is the socializing axis, related to the expanding communities of people.Huang Guowen (2002: 241-246) holds that if a functional approach is taken, discourse analysis is to analyze language in use, and study the relations between language and its context; anything related to the language in communication is a qualified research topic, for example, the choice of language forms, the relations between utterances, and cohesion, etc. This thesis mostly concerns what is context, and how context is construed and related to discourse types, so the prefix "macro" is included in the title to show the limitation of scope.In the introduction, first the terms used in the thesis are defined; then, after the research angle is proposed, this thesis puts forward its research questions; andlastly, with the introduction of the Hallidayan approach, the research procedure is established, that is, from context to text, to the scale-orientated model of discourse.And these three subtopics are discussed in detail in the following three chapters.Researches about "context" are traditionally conducted towards the categorization of those factors that affect how discourses come into being and how they are perceived. Those theories are mostly descriptive, and inefficient in their explanatory competence. In Chapter Two, "activity" is introduced as context for discourse, and activities are first divided into two kinds, discoursal and natural activities; the difference between them is not whether language is used in them or not, but lies in that the latter is the topic of the former, while the former is a component of the latter. Natural activities can be further classified into outside and inside activities. On the basis of this activity classification, the relationships between activity, discourse, and text are explored.For discourses, there are co-patternings between elements of form, content, functions and context, which are usually given the name "discourse type" or "text type" by discourse analysts. But as for how to classify discourse types, different scholars have different theories. The most commonly employed criteria are the medium of language, and rhetoric mode. By the former the distinction between spoken and written or orality and literacy is recognized; based on the latter, a typology of texts is declared to include narration, description, and argumentation. In recent years, some other influential theories are developed, including those of "generic structure potential" by Hasan (Halliday & Hasan, 1989), "taxonomy of 'discourse genres'" by Longacre (Georgakopoulou and Goutsos, 1997: 39), and the "narrative and non-narrative division" by Georgakopoulou and Goutsos (1997). But the g...