Children with autism spectrum disorder (ASD) show highly variable language skills. We are interested in what is unique about successful language learners with ASD. Statistical learning (SL), a foundational ability to gain information from patterns in the environment, is key for language acquisition in typically developing (TD) children. Importantly, unlike many measures of language competence, SL does not require social communication, which tends to be hindered in many individuals with ASD. Our large multimodal project is currently geared towards examining behavioral performance across linguistic, nonlinguistic, visual, and auditory SL tasks across children. Using safe and non-invasive brain imaging techniques, we examine the neural activity and brain signatures that underlie effective language learning. Combining findings from all of these methodologies will enable us to provide a comprehensive characterization of behavioral and brain bases of SL learning and language ability across typically developing children and children with ASD.
Word Learning and Theory of Mind
Children often require only a small amount of exposure to begin learning new word meanings when these words are presented in a supportive context. Nevertheless, figuring out what an individual word means can be a surprisingly difficult puzzle especially since children are immersed in a world of many potential referents.
Previous research has shown that children, even around the ages of 3 and 4 seem to use a range of cues derived from their communicative context (i.e. pragmatic cues), such as speaker’s intention and contrastive descriptions they hear in order to infer the meaning of new words and resolve referential ambiguities. However, little is known how learners’ ability to reason about other people’s mental states (theory of mind) relates to their success in learning and retaining inferred word meaning.
Our study, MIND, aims to address how theory of mind contributes to the immediate learning as well as long-term retention of inferred word meaning through a fun, computerized interactive word-learning paradigm.
Social Interaction and Word Learning in Autism
Social interaction is a crucial source of vocabulary input for school-aged children.Extensive research demonstrated that direct interaction (e.g., face-to-face conversations) between adults and children facilitates vocabulary development in typically developing children. Children with Autism Spectrum Disorder (ASD), who have deficits in social communication, exhibit highly variable language abilities. Thus, there is a great interest in whether direct social interaction also contributes to vocabulary learning in children with ASD. Nevertheless, research findings are mixed. Furthermore, recent studies revealed that typically developing children as young as two years old could learn nouns and verbs through overhearing adults talking even when the children were not experiencing the objects or events described in the conversations (e.g. Akhtar, 2005). However, it still remains unclear whether learning through overhearing relates to the language variability in children with ASD. Therefore, this project aims to examine word learning in multiple social contexts using fun and engaging stories. While listening to short stories, children learn words from a speaker either through direct story-telling or through overhearing. By assessing the learning outcomes using both behavioral and neuroimaging techniques, the project seeks to understand whether and how learning through direct and indirect social interaction may explain the language variability in school-aged typically developing children and children with ASD.
Prediction and Learning in Adults and Youth
Our brain uses contextual information to predict upcoming events. During language processing, the brain integrates linguistic information as sentences unfold in time and makes online predictions about upcoming words in order to process meaning efficiently. Additionally, the brain can use this prediction ability to help learn new linguistic patterns. This study aims to investigate the neural mechanism of prediction during sentence comprehension and the role of prediction in learning.
In order to do this, we combine techniques that are sensitive to millisecond changes in neural processing such as electroencephalogram (EEG) and eye-tracking. We are trying to understand neural markers of prediction and the role of prediction in learning, as well as whether prediction facilitates learning or is preceded by it.