Linguistic and visual processing of objects and actions: a coordinate-based meta-analysis of fMRI studies

Despite the literature pertaining to the neurocognitive organization of action and object concepts is now large, this topic is still controversial. The most influential theoretical perspective thus far is provided by embodied cognition theories, which are contrasted with more traditional amodal views of cognition. Evidence is especially contrasting and more limited when it comes to comparing both visually and linguistically presented stimuli, namely actions and verbs as well as objects and nouns. We used activation likelihood estimation (ALE) meta-analyses to identify similarities and differences within and between these categories. When actions, verbs, objects and nouns were considered separately, their presentation led to sensorimotor activations, providing support to embodied theories of cognition. Furthermore, linguistic and visual modalities appear to couple this sensorimotor activity with other activations related to their specific way of implementing action and object concepts. In particular, visual stimuli include a broader network of areas more linked to the actual actions and interactions with objects, while linguistic stimuli are more limited in this sense and seems to be more predominant in the left sensorimotor cortex.