A common neural code for meaning in discourse production and comprehension.

Area of Science:

• Neuroscience
• Cognitive Science
• Psycholinguistics

Background:

• Neuroimaging studies link brain activity to language semantics during comprehension.
• The neural basis of language production semantics remains less understood.

Purpose of the Study:

• To investigate how the brain codes semantic meaning during language production.
• To compare semantic coding in language production versus comprehension.
• To identify shared neural mechanisms for semantic processing in both tasks.

Main Methods:

• Functional magnetic resonance imaging (fMRI) was used to measure brain activity.
• Participants produced and listened to discourse on various topics.
• Distributional semantic models quantified speech passage similarity.
• Representational Similarity Analysis (RSA) identified neural patterns related to semantic similarity.

Main Results:

• Semantic similarity predicted neural activity patterns in a distributed, bilateral network during discourse production.
• This production network was more extensive than the comprehension network.
• Cross-task similarities revealed shared semantic coding between comprehension and production.
• Right-hemisphere regions showed semantic effects bilaterally, even when not strongly activated.

Conclusions:

• Discourse semantics engages a common neural code shared between comprehension and production.
• This shared code involves a widespread bilateral network.
• Right-hemisphere regions play a crucial supporting role in processing discourse meaning during both understanding and speaking.