| Abstract: | Study Objective: To investigate how structural variables influence readers’ construction of meaning from short-text samples of expository prose across different levels of background knowledge, text familiarity and L2 competence. Subjects: 36 Spanish-speaking medical graduates of different L2 proficiency—18 High Intermediate (HI) and 18 Advanced subjects (AD)—were randomly divided into 2 sub-groups of 9 subjects. Design: Rhetorical manipulations were performed on the published versions of three semantically different medical English abstracts (familiar and relatively unfamiliar topics). Each sub-group received one of either version and completed a reading test. A questionnaire elicited background information on the subjects. Self-generated comments on the abstracts were optional. Statistical Treatment of Data: The number of correct answers for each abstract was recorded. Between-group one- and two-way ANOVAs were applied. The questionnaires and self-generated comments were both qualitatively and quantitatively analyzed. Results: Both groups differed in their background knowledge and in their intensity of English reading. The AD subjects outperformed the HI ones in reading performance for all three abstracts together, and for the two relatively unfamiliar abstracts only. No difference was observed for the familiar abstract. In the well structured but relatively unfamiliar abstract, only L2 competence affected reading performance. In the relatively unfamiliar and poorly structured abstract, text structure, L2 competence and the interaction between both variables affected reading comprehension. Conclusions: Textual variables operate differently depending on the extent of the readers’ background knowledge and linguistic competence. Variables such as exposure to reading materials (in L1 and L2), background knowledge and L2 competence seem to outweigh the importance of structural variables. |
