How does the neural mechanism of language processing develop in children?
Saturday, October 8, 8:00 - 10:00 am EDT, Regency Ballroom
Organizer: Jin Wang1,2; 1Vanderbilt University, 2Harvard University
Chair: Matt Davis, MRC Cognition and Brain Sciences Unit, Cambridge
Presenters: Dawoon Choi, Alexis Bosseler, Alexander Enge, Jin Wang, Julie Schneider, Saloni Krishnan
Language is a complex cognitive function. Brain development supporting this higher cognitive function is prolonged, continuing over the first two decades of postnatal life. Although the neurobiology of language processing in adults has been extensively studied, how it emerges and evolves in developing children is unclear. Three general frameworks, such as maturation, skill learning, and interactive specialization account, have been hypothesized to account for human functional brain changes (Johnson, 2011). However, how these theories are supported in the domain of language processing remains to be investigated. In this symposium, our speakers will introduce their recent studies on the neural basis of language processing in both typically developing children and children with developmental language disorders. Discussions will be carried out on how studies in this field can inform the understanding of human functional brain development and suggest future educational and clinical strategies.
Sensorimotor influences on auditory speech perception in pre-babbling infants
Dawoon Choi1, Janet Werker2; 1Yale University, 2The University of British Columbia
Natural speech that infants perceive and learn from is highly multisensory. In a series of experiments, we explored whether infants’ speech perception is influenced by articulatory-auditory relations. We found behavioral (eye-tracking) and neural (EEG) evidence that preverbal infants’ auditory speech perception is influenced by articulatorily-specific sensorimotor input induced by experimentally restricting the movement of infants’ articulators. We tested perception of both native and non-native phonetic contrasts to control for the possibility of learning. To explore whether the auditory-sensorimotor relation is in place even before feedback from self-produced speech vocalizations, we tested consonant discrimination in pre-babbling infants as young as 3-month of age who are unable to produce consonant sounds. Our results show that the sensorimotor-auditory link is in place prior to specific experience watching, hearing, or producing the relevant sounds. We discuss the putative role of spontaneous and activity-dependent processes underlying this surprisingly early emerging sensitivity to articulatory-auditory relations.
Right hemisphere brain responses to familiar words at 14 months predicts infants’ future language skills
Alexis Bosseler1; 1University of Washington
Word learning is an important milestone in language acquisition and the time between 13 and 20 months of age marks a period of dramatic advances in infants’ expressive and receptive word-processing abilities. Electrophysiological event-related potentials studies indicate that during the early stages of word learning, brain activation to words bilateral, and an initial strong contribution of the right frontal brain region that attenuates as a function of age and language proficiency. We used magnetoencephalography (MEG) to explore the neural processes involved in word learning using a familiar-unfamiliar word paradigm in 14-month-olds. MEG source modeling revealed a broadly distributed network in frontal, temporal and parietal cortex that distinguished word classes between 150–900 ms after word onset. Importantly, brain activity in the right frontal cortex in response to familiar words at 14 months predicted the rate at which infants acquired new words at 18, 21, 24, and 27 months.
Meta-analyzing the cortical networks for language and semantics in children
Alexander Enge1,2, Angela Friederici3, Rasha Abdel Rahman2, Micheal Skeide1; 1Research Group Learning in Early Childhood, Max Planck Institute for Human Cognitive and Brain Sciences, 2Humboldt-Universität zu Berlin, 3Department of Neuropsychology, Max Planck Institute for Human Cognitive and Brain Sciences
Obtaining large samples remains challenging for developmental neuroimaging studies of language. This induces both false positive and false negative findings. We used coordinate based meta-analysis to identify regions that are reliably activated across fMRI studies of children’s auditory language comprehension (27 experiments, 625 children, mean age = 8.9 years, range 3–15) and semantic processing (50 experiments, 1,018 children, mean age = 10.1 years, range 4–15). The results overlapped substantially with canonical left-lateralized networks known from meta-analyses in adults. However, for language comprehension, children showed stronger effects in the right hemisphere and in the pars triangularis of the left IFG, whereas adults showed stronger effects in the pars opercularis. For semantic processing, children showed weaker effects in the left ATL. Leave-one-out and fail-safe N procedures testified robustness against publication bias. These findings provide a comprehensive picture of our current knowledge on the cortical topography of language processing during childhood.
Language specialization in typically developing children ages 5, 7, and 9 years old.
Jin Wang1,2, James Booth1; 1Vanderbilt University, 2Harvard University
One of the core features of brain maturation is functional specialization, during which some cortical regions become functionally tuned to preferred stimuli and less responsive to non-preferred ones. Successful language processing involves the coordination between phonological, semantic, and syntactic systems. Although decades of adult studies have shown specialized brain networks supporting each component during language comprehension, how children develop such specialized networks remains unclear. To address this gap, we examined phonological, semantic, and syntactic specialization at both the word and the sentence level in three groups of children ages 5, 7, and 9 years old, using functional magnetic resonance imaging (fMRI) and direct task comparisons. Our findings from multiple studies suggest that language specialization progresses from the temporal lobe to the frontal lobe as children grow older. In addition, phonological and semantic specialization appears at an earlier age than syntactic specialization.
What neural oscillations reveal about sentence processing throughout childhood?
Julie Schneider1; 1Louisiana State University
Although young children process sentences quickly and effortlessly, research indicates that the development of adult-like sentence processing abilities, and the neural structures underlying those abilities, is prolonged, continuing into adolescence (Friederici, 2005; Wang et al., 2021). This ongoing development suggests children may engage somewhat different skills than adults during sentence comprehension (Holland et al., 2007). In this presentation, I will highlight a series of empirical research studies that identify differences in the skills children engage when processing sentences using time-frequency analysis of the EEG. Across six sentence processing studies in both the auditory and written modalities, I will provide evidence that a) children in late childhood/early adolescence rely more heavily on semantics for sentence comprehension than adults, and that b) topographical and temporal changes in theta are sensitive to age differences, while changes in beta are associated with language abilities.
Examining the structure and function of frontostriatal regions in DLD
Saloni Krishnan1; 1Royal Holloway, University of London
Children with DLD struggle to learn their native language for no obvious reason. We hypothesise their language learning difficulties may be underpinned by frontostriatal abnormalities. We test this hypothesis using data from two scans from the Oxford BOLD study - the largest study examining language variation in childhood. We observed no evidence of group differences in frontostriatal regions while children performed a simple language task, verb generation (Krishnan et al., 2021, NeuroImage). Changes in this network were observed when we focused on the poorest performers, suggesting differences are attributable to task performance. We also used a novel quantitative imaging protocol that is sensitive to myelin (Krishnan et al., 2021, Biorxiv). We found that children with DLD had reduced myelin in the caudate nucleus bilaterally and across many regions in the language network. We discuss how to bring together these findings and future directions for this work.