Neural bases of image and language interactions

Abstract

Mental visual imagery and language are the two main types of representation used by humans. These two modes of mental representations are not exclusive from each other. In fact, the importance of mental imagery in human cognition comes in part from the interactions between mental image and language. Several cognitive neuroimaging studies have attempted to unravel the neural bases of these interactions. A consistent result from these studies is that mental images generated from verbal description elicit activations in both visual and language cortices. Visual cortex regions include occipital areas of the so‐called dorsal and ventral routes, whereas language areas concern both the temporal and frontal cortices, e.g., both the comprehension and production poles of language. Interestingly, it was also found that language areas are more likely to address a representation when language is the only source of information than when imagery and language are used in conjunction, which demonstrates a modulation of the activity in these areas by the mental imagery task. Overall, these findings provide a neural substrate to the dual coding theory developed by Paivio. Moreover, the fact that the same associative visual regions are activated during mental imagery triggered by either verbal description or visual encoding provides a neural base to the observation that these two types of images share common structural properties. Meanwhile, the involvement of primary visual area (PVA) during mental imagery activity remains a debated issue. First, recent results have demonstrated that PVA is not involved in some kinds of mental imagery, especially during spatial mental imagery. Second, studies have also indicated that the encoding modality, either visual or verbal, may have an impact on the neural substrates of mental images.