<p>The goal of the present study was to test the hypothesis that cross-language activation of cognates significantly increases working memory capacity for bilingual readers. Highly proficient, Spanish–English bilinguals completed the reading span task (RST) in which participants made sentence plausibility judgments while simultaneously holding target words in memory for later recall. Across three experiments the cognate status of either the constituent sentence words or follow-up targets was manipulated. Contrary to our predictions, the presence of cognates in the stimulus set did not have a uniform benefit on performance. Instead, the pattern of performance across experiments indicates that the presence of cognates caused a shift in resource allocation across the two subtasks of the RST. Specifically, when the cognate manipulation was in the target words, sentence plausibility judgments were more accurate in the cognate condition. When the cognate manipulation was in the sentences, target word recall was superior while sentence plausibility judgment accuracy was worse in the cognate condition. The ease of facilitated lexical access of cognate words may have induced this reallocation of resources to the alternative task. These findings have critical implications for the valid use of the RST as a measure of working memory capacity for comprehension in bilingual populations as well as our understanding of the working memory demands of bilingual sentence comprehension and how these are modulated by cross-language lexical activation dynamics.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

The impact of cross-language co-activation of cognates on bilingual performance on the reading span task

  • Diana Uribe,
  • Ana I. Schwartz

摘要

The goal of the present study was to test the hypothesis that cross-language activation of cognates significantly increases working memory capacity for bilingual readers. Highly proficient, Spanish–English bilinguals completed the reading span task (RST) in which participants made sentence plausibility judgments while simultaneously holding target words in memory for later recall. Across three experiments the cognate status of either the constituent sentence words or follow-up targets was manipulated. Contrary to our predictions, the presence of cognates in the stimulus set did not have a uniform benefit on performance. Instead, the pattern of performance across experiments indicates that the presence of cognates caused a shift in resource allocation across the two subtasks of the RST. Specifically, when the cognate manipulation was in the target words, sentence plausibility judgments were more accurate in the cognate condition. When the cognate manipulation was in the sentences, target word recall was superior while sentence plausibility judgment accuracy was worse in the cognate condition. The ease of facilitated lexical access of cognate words may have induced this reallocation of resources to the alternative task. These findings have critical implications for the valid use of the RST as a measure of working memory capacity for comprehension in bilingual populations as well as our understanding of the working memory demands of bilingual sentence comprehension and how these are modulated by cross-language lexical activation dynamics.