Foreign Language Comprehension: Understanding the Centrality Deficit

The goal of this study was to determine how reading in a foreign language (L2) affects one\u27s mental representation of the text and the ability to recognize and recall the text\u27s important information. Using a within-participants design, the proportion of central and peripheral ideas recalled by participants reading in their L2 was compared to that when reading in their native language (L1). Readers recalled a greater proportion of central than peripheral ideas when reading in both their L2 and L1, but when their L2 and L1 recalls were directly compared, a very interesting, yet counterintuitive, result emerged. The greatest deficit in participants\u27 L2 compared to L1 recalls was on the central, rather than peripheral, information. It is counterintuitive that the ideas that participants reading in their L2 recall the best (i.e., central ideas) are also the ideas on which they show the biggest deficit relative to their L1 recall. It is proposed that this centrality deficit stems from readers having to devote more cognitive resources to L2 word identification and consequently having fewer resources remaining to form connections among the text\u27s ideas that allow centrality to emerge. The centrality deficit was only evident among the readers who had a lower level of L2 proficiency (i.e., those for whom L2 word identification was most difficult), which supports the theory that the centrality deficit is the result of comprehending with limited resources. Additionally, having prior knowledge of the passage topic served as a compensatory mechanism for the centrality deficit. Specifically, readers with less L2 proficiency who did not have prior knowledge of the topic displayed a centrality deficit relative to their L1 recall, but this deficit dissipated when they did possess topic knowledge. This study provides insight into L2 discourse comprehension and suggests that evaluating the quality of L2 readers\u27 text recall is a worthwhile assessment that can reveal deficiencies in L2 text representations