The Relevance of Affective Information to the Semantic Representation of Social Groups

The ability to readily access social group representations and automatically categorize others as members of social groups plays a central role in our social lives. This ability helps us guide our behavior by providing a rich set of information about unknown individuals based on the existing knowledge stored in our semantic system. Neuropsychological studies reporting the presence of patients with dementia disproportionally impaired at processing social group knowledge, with spared knowledge about other categories, together with patients presenting the opposite pattern, led to propose that social groups might well be represented separately from categories such as animals, plants or objects. The organization of the semantic system in categories is consistent with several theories arguing that the relevance of the sensory and functional information, or features, forming a representation varies depending on the semantic category. The presence of a dissociation for social groups is however only accommodated by more recent theories that go beyond the simpler sensory/functional distinctions made by traditional models. Some of these theories suggest that for the representation of social groups a critical role is played by affective features. This hypothesis was corroborated by a study with brain tumor patients whose lexical semantic processing of social groups was affected by lesions of the left amygdala, insula and inferior frontal gyrus, all of which are also part of the brain network involved in processing affective information. However, beyond this finding, the empirical evidence in support of the affective features hypothesis put forward about social groups is to date relatively scarce. Thus, the aim of my dissertation is to provide the evidence of the greater relevance of affective features in social group representations by describing three original studies. First, I have reviewed the main theories of semantic memory and affect, and then I reported three original studies. In Study 1, by applying transcranial magnetic stimulation on the left inferior frontal gyrus, previously linked to processing negative affect, I showed a link between processing social group knowledge and this area. Results showed an increase in the speed at which negative social categories are categorized, without affecting responses to other stimuli. In Study 2, I reported that social category names are more susceptible to affective priming effects compared to nonsocial categories. The behavioral priming effect was also reflected in a late neural component measured via electroencephalography, where social group names displayed different amplitudes based on the affective congruence with the preceding prime. In Study 3 I documented that evaluative responses to social group names and pictures tend to be delayed compared to those towards nonsocial categories, despite no differences in categorical semantic access. Additionally, a multivariate pattern analysis (MVPA) of the neural correlates associated with this evaluation highlighted a better decoding of affective content for social groups in both an early, and a late time window across input modalities. The findings reported in my thesis provide an affirmative answer to the hypothesized greater relevance of affective features in social group representations. I argue that such relevance is expressed in terms of an affect processing region contributing to their semantic categorization, a greater benefit from affective priming and an enhanced affective decoding. These results hopefully add valuable neuroscientific evidence about how social groups are represented, contributing towards the identification of their neural substrate and affective electrophysiological correlates in terms of response magnitude and temporal dynamics.