Abstract

The current project will investigate word learning in bilingual speakers using behavioral, neuroimaging, and neurophysiological methods. Behavioral experiments will use accuracy and reaction times (RTs) as outcome measures of performance, in order to compare the processing of familiar and newly-learned words, with rich and poor semantic representations, in bilinguals’ first (L1) and second (L2) language. Participants will learn the meaning of new words (in L1 and L2) associated with many or few semantic features, over the course of 2 days. On day 3, they will be tested using semantic categorization, among other tasks. The main predictions from these studies state that lexical processing should be more accurate and faster in L1 than in L2, for familiar than newly-learned words, and for semantically rich than semantically poor items. No specific predictions are made regarding interactions between the factors. Functional magnetic resonance imaging (fMRI) will explore the neural correlates of language (L1, L2), familiarity (familiar, novel), and semantic richness (rich, poor). This experiment will require similar training for novel words to that used in behavioral experiments, so participants will learn new object names over 2 days prior to scanning. On day 3, they will perform a semantic categorization task, by pressing a button on a response box, while lying in an MRI scanner. Results from this study are expected to find stronger BOLD response in left inferior frontal gyrus (IFG) and left fusiform gyrus for L2 words versus L1 words, and novel versus familiar words. The reason for this effect is that L2 words and novel words are harder to process than L1 and familiar words, respectively, generating a more robust brain response in areas associated with executive control (IFG) and orthographic (left fusiform gyrus) processing. The comparisons familiar versus novel words and semantically rich versus poor words are expected to show greater activation in bilateral angular gyrus and precuneus in both L1 and L2. The angular gyrus plays a role in the storage of semantic representations, so it is normally more active when there is more semantic information associated with a stimulus. The precuneus seems to play a role in linking episodic memory processes with semantic processing, so it is normally active in conjunction with semantic areas. The last part of this project contemplates conducting an event-related potentials (ERPs) study in order to assess the time-course of semantic richness in bilingual word learning. After a period of training in which participants will learn new words with rich or poor semantic representations, their electrical brain activity will be recorded in order to examine N400 and N2/P2 effects. The N400 component is predicted to be modulated by language, familiarity, and semantic richness, picking at around 400 ms from stimulus onset. The P2 and N2 components have previously shown early signs of semantic processing (earlier than 200 ms) with enhanced amplitude for words with rich meaning over words with poor meaning. Hence, in the current study, P2 and N2 are also expected to elicit larger amplitude during the processing of words (familiar and novel) associated with high number of features. Results from the current project will shed light on the nature of bilingual word learning, and the representation of new word forms and semantics in the brain.