Abstract: Mirror equivalence or invariance, also known as mirror generalization or symmetry generalization, is a perceptual property in which individuals regard the left-right mirror of a visual perception object as the same stimulus. Experimental evidence of animal behavior and human cognition shows that mirror equivalence or invariance is an evolutionary adaptive process of animal and human individuals to the bilateral symmetry of natural objects, which has obvious advantages in reducing cognitive learning load and survival pressure. Recent studies on cognitive neural mechanism finds that mirror equivalence or invariance is hierarchically processed in the ventral visual pathway of primates, and the fusiform gyrus cortex in the human brain is a key region for processing mirror equivalence or invariance information of objects or faces.
Importantly, mirror equivalence or invariance may hinder the reading of script containing mirror characters, leading to mirror errors in the early reading for normal children. Therefore, it is necessary for readers to learn to use the inhibitory mechanism of “unlearning” of mirror generalization, so as to break the mirror equivalent or invariance and to acquire the ability of identifying the mirror characters. In this process, the left fusiform gyrus cortex gradually develops into the VWFA capable of recognizing mirror characters, but it still exhibits mirror equivalence or invariance for objects or faces. This is consistent with the neuronal recycling hypothesis, i.e., learning to read must occupy neurons in the left fusiform gyrus previously used for object or face processing. Furthermore, developmental dyslexia children (DD) have difficulty in inhibition of mirror generalization, suggesting the mechanism of breaking mirror equivalence or invariance may be abnormal in DD. Therefore, exploring the cognitive neural mechanism on breaking the mirror equivalent or invariance is important for elucidating the brain plasticity of learning to read.
In this paper, we first briefly discuss the evolutionary adaptive theory of mirror equivalence or invariance and early related behavioral and cognitive study evidence. Then we systematically review recent evidence on the hierarchical processing of mirror equivalence or invariance in the ventral visual pathway, the role of the fusiform gyrus cortex of human brain in the process of mirror equivalence or invariance, the cognitive neural mechanism on breaking the mirror equivalent or invariance during learning to read, and the difficulty in mirror generalization inhibition and related brain network abnormality in DD. We propose that the interaction between the left fusiform gyrus or the VWFA and the early visual cortex, the parietal cortex and the brain network of spoken language may be an important neural basis for learning to use the inhibitory mechanism of mirror generalization for breaking the mirror equivalence or invariance. Future studies are needed to focus on the role of the two hemispheres and their commissure fibers in mirror equivalence or invariance processing, the detailed processing mechanism of mirror generalization and inhibition, the influence of mirror generalization and inhibition on mirror writing, and the mirror generalization processing of Chinese characters in normal Chinese children.

Key words: mirror equivalence or invariance and its breaking, left fusiform gyrus, visual word form area (VWFA), “unlearning” mechanism, mirror generalization and inhibition