Putting the pieces together: Generating a novel representational space through deductive reasoning

Area of Science:

• Cognitive Neuroscience
• Neuroimaging
• Human Brain Function

Background:

• Representational Similarity Analysis (RSA) identifies neural bases for established representational spaces.
• Parietal activity is often observed in reasoning tasks, but its role in newly learned models is less understood.

Purpose of the Study:

• To investigate the neural implementation of a newly learned mental model derived from deductive reasoning using RSA.
• To explore the structure of parietal activity during abstract reasoning.

Main Methods:

• Participants learned relative heights through transitive reasoning, creating a linear ranking.
• Functional magnetic resonance imaging (fMRI) recorded brain activity during recall of inferred heights.
• RSA was employed to analyze brain activity patterns in relation to the learned mental model.

Main Results:

• RSA identified three right-hemisphere brain regions encoding the newly learned representational space.
• These regions include the intraparietal sulcus, precuneus, and inferior frontal gyrus.
• The findings demonstrate RSA's utility in analyzing activity patterns.

Conclusions:

• Newly learned deductive mental models are neurally implemented in specific brain regions.
• Logical reasoning appears to recruit spatial processing mechanisms, as evidenced by activity in the intraparietal sulcus.
• This study supports theories linking abstract reasoning to spatial cognition.