The role of social contextual factors in instrumental and communicative action understanding in infancy

Abstract

Humans’ ability to predict the goals of others’ actions is extensively shaped throughout life by varying social contexts and experiences. However, research has identified basic mechanisms of such an ability already in infancy. The mirror system has been implicated in this skill for goal prediction and action understanding. Similar to adults (8-13Hz), recent electroencephalography (EEG) research with infants has shown mu rhythm desynchronization (6-9Hz), which is considered the EEG signature of the mirror system functions, around the centrally located channels (corresponding to the sensorimotor cortex) for actions performed by the self and to those observed or expected from others (Cuevas et al., 2014; Fox et al., 2016; Marshall & Meltzoff, 2011). The interpretation of what the mirror system does, however, has remained equivocal. On the classic direct-matching account (Rizzolatti et al., 2001), the observation of actions directly activates an understanding of the goals qua one’s own action repertoire. The action reconstruction account (Csibra, 2008) suggests that the motor activation in the observer is only a consequence, not the cause, of predicting action goals, however, limiting this mechanism only to instrumental actions. The recently proposed social responding account (Hamilton, 2016) suggests that the mirror system activates due to one’s anticipation of an appropriate response to the observed action. Goal predictions occur in the context and service of social interaction and cooperation, which in turn, shape our predictions. Under this assumption, I conducted three EEG studies to investigate how actions and goal attribution may be interlinked with distinct social and spatial contexts that may render interpretations of actions, whether instrumental or communicative, as meaningful and goal-oriented. In studies 1 and 2, the target action that was observed by one group of adults (only in study 1) and several groups of 9-month-old infants comprised of the back-of-hand action (palm-up). The back-of-hand action has been frequently incorporated as a control condition in studies of action understanding, with the expectation that this action is not instrumental in obtaining goals (objects) when compared to more typical actions such as grasping (For example, Southgate et al., 2010). However, in the current studies, the participants watched videos of the back-of-hand action being unfolded under varying social, nonsocial and spatial contexts, as third-party observers. The findings, as evident by means of significant mu desynchronization, revealed that the back-of-hand action was interpreted as meaningful only in the congruent social setting, where the observed back-of-hand action could be associated with a requestive goal. Similarly, the pointing action, which was previously shown to not elicit mu desynchronization (Pomiechowska & Csibra, 2017), was incorporated in study 3. In this study, 12-month-old infants exhibited significant mu desynchronization when observing the pointing action in social situations, either for a requestive purpose or to share attention. The significant findings in this thesis, particularly from the infant samples, were elicited in the right hemisphere, i.e., the right centro-parietal and the right fronto-central regions, providing further support to the social functions of the right hemisphere. Further, the findings provided substantial evidence that processing communicative actions and attributing communicative goals also exhibited mu rhythm desynchronization, suggesting that the function of the mirror system is not limited to instrumental actions alone. Finally, the findings also bring to light the limitations of existing theories of action understanding and suggest that a broader, composite perspective may be better suited that offers more flexibility for one’s interpretation of the association between actions, context and goals.