The domain-general and domain-specific role of the semantic neural network in mathematical processing

Neuroimage. 2024 Dec 20:120985. doi: 10.1016/j.neuroimage.2024.120985. Online ahead of print.

Abstract

The role of the visuospatial network in mathematical processing has been established, but the role of the semantic network in mathematical processing remains poorly understood. The current study compared different types of inductive reasoning with the functional magnetic resonance imaging (fMRI) technique to investigate the role of the semantic network in mathematical processing and whether the role is domain-general or domain-specific. 32 undergraduate students were recruited to complete tasks involving numerical, geometrical, situational, and verbal inductive reasoning, as well as arithmetical computation. The intensity and pattern analysis of brain activation found that all types of inductive reasoning elicited greater and similar activation than arithmetical computation in the semantic network, mainly including the left middle temporal gyrus, inferior frontal gyrus, angular gyrus, and dorsomedial prefrontal cortex. Furthermore, the intensity and patterns analysis of functional connectivity found that numerical inductive reasoning elicited stronger and dissimilar connectivity between the semantic and visuospatial networks than other types of inductive reasoning and arithmetical computation. These findings suggest that the semantic network not only plays a domain-general role but also exhibits domain-specific interactions with the visuospatial network in mathematical processing.

Keywords: domain-general; domain-specific; functional magnetic resonance imaging; inductive reasoning; mathematical processing; semantic network.