Abstract rule generalization for composing novel meaning recruits a frontoparietal control network.

Area of Science:

• Neuroscience
• Cognitive Science
• Linguistics

Background:

• Cognitive flexibility and language generalization are vital for human adaptation.
• Inferring meanings of novel compositional words, like "un-reject-able-ish", highlights language's combinatorial nature.
• The neural underpinnings of this compositional ability are not fully understood.

Purpose of the Study:

• To investigate the neural mechanisms and representations involved in deriving meaning from novel compositional words.
• To explore how abstract relational structure rules influence word meaning based on sequential affix-stem order.
• To identify brain networks engaged during semantic priming with novel compositional words.

Main Methods:

• Functional magnetic resonance imaging (fMRI) was used to monitor brain activity.
• Participants learned a semi-artificial language with compositional words governed by abstract rules.
• A semantic priming task involved novel compositional words (primes) and real word targets (synonyms).

Main Results:

• Compositional processes activated a widespread temporoparietal network.
• Representations of composed word meanings were found in a left-lateralized language network.
• Newly composed meanings were decodable from brain activity during prime presentation, independent of prime word representations.
• Abstract rule representations for composition engaged a bilateral frontoparietal network.

Conclusions:

• Language composition and generalization rely on a bilateral frontoparietal circuitry for abstract rule application.
• Neural representations of newly composed word meanings are distinct from those of their constituent parts.
• The findings advance our understanding of the neural basis of semantic composition and cognitive flexibility in language.