摘要

Sensorimotor representations in the brain encode the sensory and motor aspects of one%26apos;s own bodily activity. It is highly debated whether sensorimotor representations are the core basis for the representation of action-related knowledge and, in particular, action words, such as verbs. In this review, we will address this question by bringing to bear insights from the study of brain-damaged patients exhibiting language disorders and from the study of the mechanisms for language acquisition in infants. Cognitive neuropsychology studies have assessed how damage to representations supporting action production impacts patients%26apos; ability to process action-related words. While correlations between verbal and nonverbal (motor) impairments are very common in patients, damage to the representations for action production can leave the ability to understand action-words unaffected: likewise, actions can still be produced successfully in cases of impaired action-word understanding. Studies with infants have evaluated the relevance of sensorimotor information when infants learn to map a novel word onto an action that they are performing or perceiving. These results demonstrate that sensorimotor information is insufficient to fully account for the complexity of verb learning: in this process, infants seem to privilege abstract constructs such as goal, intentionality and causality, as well as syntactic constraints, over the perceptual and motor dimensions of an action. Altogether, the empirical data suggest that, while not crucial for verb learning and understanding, sensorimotor processes can contribute to solving the problem of symbol grounding and/or serve as a primary mechanism in social cognition, to learn about others%26apos; goals and intentions. By assessing the relevance of sensorimotor representations in the way action-related words are acquired and represented, we aim to provide a useful set of criteria for testing specific predictions made by different theories of concepts.

  • 出版日期2012-6