Serial Order in Language Learning in Bilinguals

Abstract

The current thesis has two aims to further the understanding of the cognitive underpinnings that are involved in word-learning and language processing. One aim is to understand how individuals are able to make temporary phonological and serial order representations of new words in language and non-language domains. The second aim is to investigate whether the mechanisms involved in maintaining temporary nonverbal and serial order representations are related to verbal short-term abilities. We created four behavioural tasks to determine the processing of phonological short-term memory information, nonverbal short-term memory information, serial order short-term memory information and rhythmic short-term memory. We used 30 adult Spanish-English bilinguals as the target population to also investigate whether individual language abilities in two spoken languages affect the learning of words in a foreign language with distinct prosody and phonology. The first correlational analysis revealed that performance on a serially ordered verbal short-term memory task that involved a language of unfamiliar prosody and phonology was strongly predicted by performance on two serially ordered verbal short-term memory tasks that involved languages of familiar prosody and phonology. A second correlational analysis showed that tasks that tapped into individuals’ memory for serial order in the verbal, nonverbal and rhythmic domains were weakly associated with one another. In a third correlational analysis, it was shown that individuals’ lexical knowledge of Spanish was not a predictor of their performance on a measure of their serially ordered verbal short-term memory abilities. Multiple linear regression analyses found that none of the tasks that were used to measure individuals’ abilities for processing serial order information in the verbal, nonverbal and rhythmic domain were strong predictors for foreign-word learning. Overall, the results show promising findings for the tasks that tapped into serial order short-term memory for verbal information. However, they also suggest that the nonverbal and rhythm tasks may not be reliable measures of the constructs we were hoping to study. Future work should adjust the tasks to ensure we are properly tapping into individuals’ serial order abilities in the nonverbal and rhythm domains.