Presentation

Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Poster Slams

Shared Neural Correlates Underpin Theory of Mind and Semantic Cognition: Evidence from a Meta-Analysis of 344 Functional Neuroimaging Studies

Poster E7 in Poster Session E, Saturday, October 8, 3:15 - 5:00 pm EDT, Millennium Hall

Eva Balgova¹, Veronica Diveica¹, Rebecca L. Jackson², Richard Binney¹; ¹School of Human and Behavioural Sciences, Bangor University, Gwynedd, Wales, ²MRC Cognition and Brain Sciences Unit, University of Cambridge, UK

Currently, there is a lack of consensus regarding whether brain regions involved in social processing are specialised for that domain, and to what extent domain-general systems contribute to social cognition. Recently, it has been proposed that the semantic system involved in extracting meaning from language is also crucial for the retrieval of meaning from other types of social cues. However, there is limited direct evidence for a role of semantic processing in social cognition. Therefore, the present study aimed to elucidate the extent to which key nodes of the semantic cognition (SC) network are involved during social mental state inference or theory of mind (ToM) tasks. To overcome the limitations of single studies that often lack power and are prone to effects of idiosyncratic designs and analytic choices, we adopted a meta-analytic approach. We used the activation likelihood estimation algorithm to synthesise the results of a large set of functional neuroimaging data. Our aim was to a) establish the degree of overlap between the ToM network and a wider SC network and b) investigate the influence of methodological and task design factors, such as stimulus domain (verbal/non-verbal) and sensory input modality (visual/auditory) on network overlap. This allowed us to test domain-generality/specificity and whether the pattern of overlap could be explained by the type of tasks/stimuli. To address our main aim, we manually extracted two sets of activation coordinates from 133 ToM, and 211 SC experiments and these were subjected to formal conjunction and contrasts analyses. To address our secondary aim, we repeated these analyses on data subsets that were split according to stimulus domain and input modality. The results revealed extensive overlap between the brain activation reliably associated with ToM and that correlated with SC. The overlap included several key semantic brain areas: the anterior temporal lobes, the angular gyrus (extending to the intraparietal lobule and temporo-parietal junction (TPJ)), the posterior middle temporal gyrus and the inferior frontal gyrus. Further, after the data split, the conjunction pattern remained consistent for the verbal and visual experiments. Although the conjunction between non-verbal ToM and SC experiments was diminished in brain areas that are part of the semantic representational system, we still observed overlap in key nodes of the semantic control network. The right TPJ region consistently activated for ToM but not for SC. To our knowledge, this is the first attempt to formally assess the topological overlap of network activated by ToM and SC across an extensive range of semantic and social studies. The finding that ToM and SC share neural correlates supports the claim that ToM may draw on cognitive processes related to semantic retrieval and speaks against the view that ToM relies solely on domain-specific social neurocognitive systems. Instead, we add to the rising body of evidence supporting a role for the semantic network in social cognition. Key Words: semantic cognition, social cognition, social semantics, ALE meta-analysis

Topic Areas: Meaning: Combinatorial Semantics, Methods