Theoretical assumptions in L2 reading research

Only recently has research on reading in Japanese as a second language (JSL) started to appear in major journals in the fields of applied linguistics and second/foreign-language learning and teaching. Research on second-language (L2) reading has developed through the application of ideas from first-language (L1) reading research, mainly in English. Drawing on L1 models, L2 reading has been conceptualized as an active process in which the reader constructs meaning through interaction with a text, and a dynamic process in which bottom-up (text-based, data-driven) processing and top-down (knowledge-based, conceptually driven) processing take place interactively (Bernhardt, 1991; Swaffar, Arens, & Byrnes, 1991). Furthermore, this process is constrained by the available processing capacity.

Although there are currently a number of L1 reading models (e.g. Gernsbacher, 1994), most share the following assumptions about the cognitive process. The reader identifies letters and characters, recognizes words, and analyzes the syntactic and semantic relations in sentences to extract propositions. These propositions are integrated with information activated from general knowledge (i.e. inference) and encoded into a coherent representation of the text in memory. The resulting memory representation consists of multiple levels, including at least surface code, propositional text-base, and situation model (van Dijk & Kintsch, 1983). Throughout the course of processing a text, the reader monitors and regulates processing according to the comprehension goal in a given situation.

The differences between reading in the L1 and the L2

L1 reading research has a profound influence on L2 reading research. However, it is also important to note that, for all the underlying similarities between the two, L2 reading is different from L1 reading in a number of significant respects.