Presentation

Elaina S Jahanfard¹, Tamara Y Swaab¹; ¹University of California, Davis

There is abundant evidence to suggest that predictive processing facilitates language comprehension in monolinguals by pre-activating linguistic representations for upcoming language input (e.g., Kuperberg & Yeager, 2016). Despite the global ubiquity of bilingualism, studies that examine the type of and extent to which features of words are being pre-activated when bilinguals process their second language (L2) are more limited. Previous work suggests that Spanish-English bilinguals show greater facilitation of information at the word form level than monolinguals do (Hoshino & Kroll, 2008). But these studies did not address whether these facilitatory effects can be attributed to pre-activation of word form information in the L2 of Spanish-English bilinguals prior to the presentation of the target word. The proposed study aims to address this question by examining prediction during word processing in Spanish-English bilinguals. 50 Spanish-English bilinguals and 50 English monolinguals will be recruited to participate in a two word priming study with a prediction task. Materials from Gao et al. (2022) will be normed within a different sample of 60 Spanish-English bilinguals to establish that the forward association between primes and targets is the same in this population. Before the norming study and the experiment, all bilingual participants will complete a language history questionnaire (LHQ3) and a validated test of English proficiency (LexTALE) to establish their language proficiency of L2 (Li et al, 2020; Lemhöfer & Broersma, 2012). In the priming study, participants will be presented with 480 prime-target word pairs. Target words are semantically related (circus-clown) or unrelated to the primes (trim-clown). Their task is to predict the target words and to produce (name) the predicted word prior to the presentation of the target (within 1800ms). They then make a lexical decision to the target words. The pre-target naming response will be compared to the presented visual target word to assess whether it is identical (accurate prediction), semantically related or unrelated, and lexical decision latencies will be analyzed accordingly. We will replicate this study using ERPs. Previously, we have found that monolingual participants showed facilitated processing of accurately predicted target words relative to semantically related target words, which indicates processing benefits of lexical pre-activation (Gao et al, 2022). If bilinguals pre-activate lexical form in their L2, they should show this same pattern of results. Preliminary norming results from 18 proficient Spanish-English bilinguals replicated the 50% forward association strength that was found in a monolingual norming study with 60 participants. LMER analyses will be performed and we predict main effects of prediction accuracy and relatedness on RT latency and the N400 event-related potential (ERP) component. An interaction of prediction accuracy by relatedness would be consistent with pre-activation of word forms in L2. This finding would indicate that bilingual language users selectively activate words in the context language when they are engaged in an active prediction task.

Topic Areas: Multilingualism, Language Production