摘要:Historically, there has been an intense debate over the most effective approach to instruction: bottom-up versus top down (Adams, 1990; Chall, 1967; Grabe, 1991; McCarthy, 1991; Swaffar, 1988). With the increasing emphasis on the role of context in literacy activities (Anderson and Pearson, 1984; Lave, 1988; Steffensen, Joag-dev, and Anderson, 1979), reading research in English as a second language (ESL) has in the recent past focused on the effects of prior knowledge, or schemata, on comprehension (Carrell, 1984, 1987; Dubin and Bycina, 1991; Lee, 1986). Most of these studies generally reported significant effects of background knowledge on reading performance. Concomitantly, the proliferation of recent instructional approaches also reflected a clear tendency to over emphasize the role of background knowledge and devalue the role linguistic knowledge plays in the comprehension process (Fang, 1993; Yule, 1986). Mitchell (1982) and Perfetti (1989) pointed out that the experimental manipulations in such studies often tilted the balance in favor of the kind of processing that is guided by top-down effects. Therefore, caution needs to be exercised in the interpretation of their findings.