The Origin of Cognitive Modules for Face Processing: A Computational Evolutionary Perspective

Abstract

Despite extensive research, understanding how cognitive modules emerge remains elusive due to the complex interplay of genetic, developmental, and environmental factors. Computational modeling, however, provides a means of exploring their origins by directly manipulating these factors. Here we aimed to investigate the emergence of cognitive modules by developing the Dual-Task Meta-Learning Partitioned (DAMP) model, whose plastic architecture facilitates automatic structure optimization through a genetic algorithm that emulates natural selection by iteratively selecting for efficient learning fitness. We found that a specialized module for face identification robustly emerged in the DAMP model. Critically, the emergence of the face module was not influenced by the demands of cognitive tasks (identification versus categorization) or the type of stimuli (faces versus non-face objects). Instead, it was determined by the structural constraint of sparse connectivity within the network, suggesting that the face module may arise as an adaptation strategy to challenges posed by sparse connections in neural networks, rather than being an information processor required by certain stimuli or tasks. These findings provide a new evolutionary perspective on the formation of cognitive modules in the human brain, highlighting the pivotal role of the structural properties of neural networks in shaping their cognitive functionality.