The language network responds robustly to sentences across tasks

Area of Science:

• Neuroscience
• Cognitive Science
• Psycholinguistics

Background:

• Language processing relies on a network of left frontal and temporal brain regions.
• The role of task context in modulating neural responses within language networks is not fully understood.
• The domain-general multiple demand (MD) network is involved in cognitive control and task demands.

Purpose of the Study:

• To investigate the stability of language network responses across various task contexts.
• To compare the task sensitivity of the language network with the domain-general MD network.
• To determine how task demands influence linguistic computations and neural network recruitment.

Main Methods:

• Functional magnetic resonance imaging (fMRI) was used to measure brain activity in 52 participants.
• Participants performed language tasks (reading sentences and nonword lists) under six different conditions, including passive reading and semantic engagement.
• Multivariate analyses were employed to decode stimulus and task information within the language and MD networks.

Main Results:

• Language-responsive brain regions (language network) showed stable responses to sentences across all tasks, with participant-specific localization.
• The MD network responded to both sentences and nonwords when a task was present, indicating task sensitivity.
• Stimulus information was more decodable in the language network, while task information was more decodable in the MD network.

Conclusions:

• Language processing is robust to variations in task context, with a dedicated language network.
• The language and MD networks have complementary functions: the language network extracts linguistic information, and the MD network guides task-specific responses.
• Care must be taken when designing language paradigms, as task demands can engage both independent networks.