Neural correlates of dependency length processing during natural language comprehension.

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Linguistics

Background:

• Extensive research exists on syntactic processing neural mechanisms.
• The brain's processing of dependency length, a key syntactic feature, remains under-investigated in natural language comprehension.

Purpose of the Study:

• To systematically investigate the whole-brain neural correlates of dependency length processing.
• To develop and apply a novel syntactic parsing framework for neuroimaging data.

Main Methods:

• A two-stage progressive syntactic parsing framework was developed.
• Functional magnetic resonance imaging (fMRI) data from natural narrative listening tasks were analyzed.
• Neural activation patterns related to dependency length were identified.

Main Results:

• Neural correlates for dependency length processing were consistently identified across individuals.
• Sensitive brain regions were distributed across the frontal-temporal-parietal language network, indicating multi-region collaboration.
• Differences in neural activation similarity between syntactic structures and dependency length were observed, potentially linked to variability and integration needs.

Conclusions:

• Dependency length processing engages a distributed brain network, supporting the dependency length minimization hypothesis.
• This study provides a method for localizing fine-grained syntactic processing and reveals associations between dependency length and syntactic structures.
• Advances understanding of the neurobiology of abstract syntactic processing.