Neural correlates of social perception: The posterior superior temporal sulcus is modulated by action rationality, but not animacy

Neural correlates of social perception: The posterior superior temporal sulcus is modulated by action rationality, but not animacy Ben Deen (bdeen@mit.edu) and Rebecca R. Saxe (saxe@mit.edu) Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology Cambridge, MA 02139 Abstract Pelphrey, Morris, & McCarthy, 2004; Pelphrey, Singerman, Allison, & McCarthy, 2003; Saxe, Xiao, Kovacs, Perrett, & Kanwisher, 2004; Vander Wyk, Hudac, Carter, Sobel, & Pelphrey, 2009). Such actions have been referred to as incongruent, irrational, or unexpected. This effect has been interpreted as evidence that the pSTS is sensitive to the goals or intentions underlying human motion. For instance, Pelphrey et al. (2004) argued that the pSTS is involved in predicting actions in a given context based on an “intentional stance,” in which actions are determined by a goal state and an assumption that the agent will choose the most efficient means to achieve the goal given situational constraints. They proposed that when this prediction is violated, the pSTS must engage in extra processing to explain the observed action in other terms, which would explain its stronger response to unexpected actions. Another line of research supporting the role of pSTS in action understanding as employed animations of simple geometric shapes as stimuli (Castelli, Happe, Frith, & Frith, 2003; Gobbini, Koralek, Bryan, Montgomery, & Haxby, 2007). These studies have found a stronger pSTS response to animations depicting social interactions between animate shapes, compared with animations of shapes moving as inanimate physical objects. This demonstrates that the role of the pSTS extends to animations that lack the form and motion kinematics of humans, but imply intentional action. However, such comparisons have been largely visually uncontrolled, and could also reflect one of a number of processes: detecting agents, processing of their motion or intentions, or processing of interactions between multiple agents. The present study aimed to investigate the neural correlates of social perceptual processes, using geometric shape stimuli. In particular, we use dot-chain stimuli perceived as slithering snakes or worms, which provide a strong percept of animacy without the need for multiple, interacting agents (Gao, New, & Scholl, 2011). This ensures that any effects observed do not relate to processing interactions between agents (c.f. Centelles, Assaiante, Nazarian, Anton, & Schmitz, 2011). To investigate each of the subprocesses listed above, we separately manipulated the perceived animacy, goal-directedness, and path rationality (or expectedness) of the animations. We first performed a behavioral study, eliciting judgments about these animations on various dimensions. The animations were then used as stimuli for an fMRI experiment, to investigate the response of the pSTS, as well as motion- sensitive area MT+, as a control region. Recent research has investigated the neural basis of social perception, the ability to make high-level social inferences from perceptual information. The right posterior superior temporal sulcus (pSTS) has been identified as a candidate region for this ability, but the specific processes to which the pSTS contributes remain unclear. In the present study, we investigated the neural correlates of social perception using simple animated geometric shape stimuli, separately manipulating the perceived animacy, goal-directedness, and path rationality in the animations. We did not find an increased pSTS response to animate or goal-directed animations. However, we found that across conditions, the pSTS response tracked path rationality, with stronger responses to irrational paths. This is consistent with prior neuroimaging research on the perception of human actions, and supports the claim that the pSTS is involved in action understanding. Keywords: social perception, fMRI, superior temporal sulcus Introduction Humans have a remarkable ability to infer the dispositions and intentions of other agents from perceptual information, and specifically from motion patterns such as hand and body motion, gaze shifts, and facial motion. This ability, termed social perception, comprises a number of subprocesses: the detection of agents in an environment, perceptual analysis of their motion, inference about social properties from the agent’s actions and their context, and prediction of future actions based on these properties. Recent research has begun to probe the neural basis of these processes, although the relevant brain regions and their specific functional role is still debated. One line of research has pointed to the right posterior superior temporal sulcus (pSTS) as a critical region for social perception (Allison, Puce, & McCarthy, 2000). This region responds more strongly to (human) biological motion than motion of inanimate objects (e.g. Grossman et al., 2000; Pelphrey et al., 2003). These responses might relate to the detection or perceptual analysis of biological motion, to higher-level processing of the intentions underlying the actions, or to some combination thereof. Another set of studies indicates that the pSTS response to human actions is modulated by inferred intentions. Specifically, actions that violate inferred intentions in a given context, such as twisting empty space next to a gear rather than a gear itself, elicit a stronger pSTS response than the expected actions, across a range of contexts and specific actions (Brass, Schmitt, Spengler, & Gergely, 2007;