This study examined relationships among measures of language proficiency, cognitive style, and metaphor comprehension. Subjects were university students who were native English speakers or who were enrolled in a course on English as a second language (ESL). Consistent with predictions, native English speakers scored better than ESL students on academic measures of English proficiency, but there were no group differences on level of cognitive sophistication in English metaphor interpretation or on a measure of metaphor fluency (number of metaphor interpretations produced). For ESL students, metaphor fluency was positively related to a measure of English communicative proficiency, whereas a measure of field independence was negatively related with both metaphor fluency and communicative proficiency. These findings on cognitive style are consistent with theoretical predictions that have heretofore found little empirical support in the second language literature.A question of enduring interest to second language researchers concerns factors that influence the acquisition of a second language (e.g., Naiman, Frolich, Stern, & Todesco, 1978;Skehan, 1989). Cummins (1991) made a distinction between attribute-based and input-based aspects of second language proficiency. The acquisition of attribute-based aspects of proficiency is heavily influenced by stable attributes of the learner, for example, cognitive and personality variables. The acquisition of input-based aspects is related to the quality and quantity of second language input in the environment.In his investigation of attribute-based aspects of second language proficiency, Cummins made the further distinction between decontextualized versus contextualized language use (Snow, 1987). This dimension of language use varies according to the amount of contextual support for expressing and receiving meaning. In contextualized use, the participants can actively negotiate meaning, and the language is supported by situational cues. Decontextualized use relies primarily on linguistic cues to meaning in the absence of supporting context. Cummins reviewed data on the interdepen-

No abstract

No abstract