202510.24
How the brain learns meaning-based grammatical rules (Publication)
Posted in RESEARCH
How does the brain learn grammatical rules that are grounded in meaning rather than form?
To address this question, we conducted an fMRI study to examine how people learn nominal classification rules—grammatical systems that categorize nouns according to semantic features such as animacy and object size. Thirty-six Japanese native speakers learned a semi-artificial language containing three noun classes (animate, small inanimate, large inanimate) while their brain activity was recorded. Results revealed that activation in the left middle temporal gyrus (MTG) predicted successful learning and rule generalization, and that activity in the left inferior frontal gyrus (IFG) and caudate nucleus decreased as learning progressed, indicating their role as part of the core network for grammatical learning. Participants with higher working memory capacity exhibited reduced thalamic activation, suggesting more efficient information processing. These findings demonstrate that meaning-based grammar learning relies not only on formal rule processing but also on the cooperation of broad neural networks responsible for semantic and procedural learning. This study was published in Brain and Language.(Jeong)
🧠 “The neural correlates of nominal classification rule learning and their individual differences” Diego Elisandro Dardon, Haining Cui, Motoaki Sugiura, and Hyeonjeong Jeong, Brain and Language
🔗 https://doi.org/10.1016/j.bandl.2025.105654
How does the brain learn grammatical rules that are grounded in meaning rather than form?
To address this question, we conducted an fMRI study to examine how people learn nominal classification rules—grammatical systems that categorize nouns according to semantic features such as animacy and object size. Thirty-six Japanese native speakers learned a semi-artificial language containing three noun classes (animate, small inanimate, large inanimate) while their brain activity was recorded. Results revealed that activation in the left middle temporal gyrus (MTG) predicted successful learning and rule generalization, and that activity in the left inferior frontal gyrus (IFG) and caudate nucleus decreased as learning progressed, indicating their role as part of the core network for grammatical learning. Participants with higher working memory capacity exhibited reduced thalamic activation, suggesting more efficient information processing. These findings demonstrate that meaning-based grammar learning relies not only on formal rule processing but also on the cooperation of broad neural networks responsible for semantic and procedural learning. This study was published in Brain and Language.(Jeong)
🧠 “The neural correlates of nominal classification rule learning and their individual differences” Diego Elisandro Dardon, Haining Cui, Motoaki Sugiura, and Hyeonjeong Jeong, Brain and Language
🔗 https://doi.org/10.1016/j.bandl.2025.105654