Search Abstracts | Symposia | Slide Sessions | Poster Sessions
The Influence of Modality and Early Language Experience on the Neural Network for Language
Poster Session E, Sunday, September 14, 11:00 am - 12:30 pm, Field House
Brennan Terhune-Cotter1,2, Laurie Glezer11, Stephen McCullough1, Sarah Kimbley1, Zella Pangle1, Sarah Kimbley1, Karen Emmorey1; 1San Diego State University, 2University of California, San Diego
This study investigates the neural correlates of language comprehension in a sample of deaf adults who are bilingual in American Sign Language (ASL) and written English, and who experienced variable language access in early childhood. We obtained functional MRI data from 20 deaf signers with ages of ASL exposure (AoE) ranging from birth to 20 years (mean AoE = 8.3; SD = 7.1 years). Our sample is characterized by relatively high proficiency in both ASL and English, despite delayed early language access. Participants viewed sentences and pseudo-sentences in written English and in ASL, along with a visual baseline condition for ASL. Pseudo-sentences in each language were sentences made up of phonologically legal but non-existing words or signs. We designed pseudo-sentences to elicit equivalent phonological processing but less syntactic/semantic processing when compared to sentences. We derived effect sizes for all conditions by estimating responses within subject-specific functional regions of interest (fROIs), which were defined using a contrast between ASL sentences and the visual baseline. Our fROIs were constrained to homologous regions known to be involved in ASL comprehension: frontal regions (IFG and MFG), temporal regions (posterior and anterior temporal cortex), and occipitotemporal regions (including fusiform cortex). We evaluated the effects of language (ASL vs. English) and stimulus type (sentences vs. pseudo-sentences) on neural effect sizes using a linear mixed-effects model. Broadly, ASL elicited stronger activation as compared to English, particularly within occipitotemporal and posterior temporal regions. Neural activation was bilateral with a slight bias to the right hemisphere for both ASL and English; ASL elicited stronger right-lateralization compared to English. The comparison between sentences and pseudo-sentences did not differ across languages, with sentences eliciting greater activation than pseudo-sentences in bilateral frontal and temporal regions, but not in occipitotemporal or fusiform regions. To investigate the effect of early language experience on neural activation, we included ASL AoE as a covariate in the model. There were two interaction effects: Participants exposed to ASL later had more right-lateralized activation for English, and they also had stronger activation in occipitotemporal regions for ASL. There was a small effect of AoE in the MFG, in which delayed ASL exposure led to attenuated activation for ASL, but AoE did not modulate activation in the IFG. AoE did not interact with the type of input (pseudo-sentences or sentences), suggesting that effects of delayed early language occur at the phonological level in this population. Our results suggest that the right hemisphere is recruited for both ASL and written English comprehension in deaf bilinguals. Additionally, delayed language exposure predicts a rightward shift in activation for written English. Bilateral activation for both written English and ASL comprehension has been observed in previous studies (Neville at al., 1998), and strong right-hemisphere activation has been documented in late signers (Ferjan Ramirez et al., 2014; 2016; Mayberry et al., 2018). Our findings replicate prior research showing that a delayed AoE increases occipitotemporal activation, but we do not find attenuated frontotemporal activation with later language exposure (Twomey et al., 2020; Mayberry et al., 2011).
Topic Areas: Signed Language and Gesture, Multilingualism