Speaker
Description
Long-term memory is conventionally separated into distinct, interacting subsystems: declarative (DM) and procedural memory (PM). The declarative-procedural (DP) model of language learning posits a neurocognitive shift from reliance on explicit DM retrieval to more efficient, automated PM retrieval for language processing as mastery is achieved, particularly for rule-based grammatical content. However, the trajectory of precise roles of each memory subsystem during the development of grammatical processing in childhood is poorly understood. Additionally, it is unknown whether individual differences in the trajectories of language development impact the extent of engagement of each system for early grammatical processing.
This study leverages functional magnetic resonance imaging (fMRI) data collected in 7-year old children and multilevel modeling to quantify neural response patterns within established DM/PM brain regions during an auditory grammar task. We assess whether, based on the DP model, greater PM engagement corresponds with higher accuracy in grammar tasks and whether the extent to which this effect varies among same-aged children.
Our findings reveal robust, systematic differences in both DM and PM activity when processing grammatical errors compared to grammatically correct auditory sentences as well as significant subject-level variability in the effects of task demand on brain activity. We also identify a positive impact of accuracy on left PM engagement, but only during the most challenging grammatical error task. Our analyses introduce further insight into the unique roles of DM and PM memory systems during language learning, while also highlighting the impact of individual differences in children’s neurocognitive development during the language learning processes.
