The case for interactive models of reading

Anyone concerned with ESL reading cannot help but be struck by the dramatic improvements in ESL reading theory and practice during the past ten years or so. These improvements may be traced to multiple sources, including, for example, the general movement toward more communicative kinds of language teaching and a new concern for needs analyses in relation to particular populations of learners. But clearly the major source of improvement has been the growing understanding and acceptance of psycholinguistic models of the reading process, especially as represented in the work of Kenneth Goodman and Frank Smith. Proceeding from the views of Goodman and Smith, such specialists as Eskey (1973), Clarke and Silberstein (1977), Coady (1979), Carrell and Eisterhold (1983; reprinted as Chapter 5 in this volume) and Carrell (1983) – and of course many others – have adapted this socalled top-down approach to second language reading, and have tried to relate it to the practical problems of curriculum design and teaching methods and materials.

Terms like top-down or the contrasting bottom-up (or, for that matter, interactive) are, of course, merely metaphors for the complex mental process of reading, top here referring to such “higher” order mental concepts as the knowledge and expectations of the reader, and bottom to the physical text on the page. Proponents of each of these approaches have argued that the former or the latter is the true starting point and thus the controlling factor in the process.