On the structure of scientific texts

1. In the study of text structure there are basically two approaches which we may adopt. On the one hand, we may seek to understand the mechanisms of textual cohesion, to discover what makes a sequence of sentences into a coherent text. Such a study involves the discussion of anaphora, of reference and substitution, of ellipsis, of the role of conjunctions and ‘sentence adverbs’, of lexical and semantic cohesion between sentences, and so forth. The alternative approach is to take a more global perspective, to seek to understand the overall organisation of texts, to understand how one episode of a narrative develops from another and how paragraphs and chapters are built into cohesive wholes. Following van Dijk (1972) we may call the first approach the study of ‘micro-structure ’ and the second the study of ‘macro-structure’. The latter has long been the province of the literary critic, the rhetorician, and the analyst of folk-tales (e.g. Kinneavy 1971, Propp 1968). It is an area which linguists have neglected. Indeed it is only relatively recently that they have turned their attention to text structures at all, and it is perhaps natural that they should have concentrated primarily on ‘micro-structure ’ (e.g. Grimes 1975, Halliday and Hasan 1976) since it is here that techniques used in the analysis of sentences can be most easily applied