By what process do individuals acquire a second language? During early stages of learning second language words, people often report being aware of their first language. However, as they become more proficient in the second language, they begin to report that they can think directly in the second language without reference to the first language. Although there is empirical evidence to support these common observations, we do not fully understand how memory changes to accommodate the acquisition of words in a new language, or how skilled bilinguals develop cognitive mechanisms that allow them rapid access to each language and also to the interconnections between their languages. The general goal of this research is to explore the changes in memory that occur as adults acquire a second language, and to compare the performance of beginning second language learners to skilled bilinguals. A specific goal of the research is to test and evaluate a new model of cross-language interconnection in which the connections between the two languages are thought to be asymmetric. The model proposes that there are asymmetries in the strength of the connections between lexical (word-to-word) representations in different languages. The model also proposes asymmetries between lexical representations and concepts. Two lines of research will test the predictions of the new model. The first concerns predictions for cross-language performance in naming, translation, priming, and memory tasks. A central claim in this work is that there are two different routes to translation: translation from the first language to the second language requires understanding the meaning of the second language word, whereas translation from the second language to the first can be accomplished on the basis of associations between words in the two languages. One set of experiments will evaluate the evidence for this claim and also test implications for later memory performance. A second line of research will explore the development of the hypothesized cross-language asymmetries over the course of second language learning in adults by determining the factors that influence how quickly an individual becomes fluent in the second language. The results of the proposed experiments will have important implications for theories of memory and for understanding levels of language processing and their interaction. In addition, by identifying the way in which a new memory system is established for second language words during second language acquisition, the research will suggest ways in which basic cognitive research can be applied to second language instruction and bilingual education. Given the changing demographic characteristics of our society, this is an issue of increasing importance.
