Just as important as what we say is how we act, which critically impacts our understanding of each other. Nonverbal social cues, including facial expressions and body movements, are crucial to interpersonal competence and

the regulation of social interactions. Autism is a neurodevelopmental disorder whose symptom cluster includes unique impairments in communication and social interaction (Hobson, 1986), resulting in a dearth of adaptive social behavior. Little is known about the pathobiology supporting aberrant social behavior in autism, although converging evidence from several recent studies suggests that cortex-wide abnormal functional connectivity patterns are more influential than local deficits in specific regions (Belmonte et al., 2004; Courchesne & Pierce, 2005; Wickelgren, 2005). Studies finding reduced connectivity in autism include those investigating language processing (Just, Cherkassky, Keller, & Minshew, 2004), visuomotor skills (Villalobos, Mizuno, Dahl, Kemmotsu, & Müller, 2005), inhibition (Kana, Keller, Minshew, & Just, 2007), Theory of Mind (Castelli, Frith, Happé, & Frith, 2002), and fixation resting state (Cherkassky, Kana, Keller, & Just, 2006). In addition, several recent studies of white matter structure have reported disordered cortico-cortical connections in autism (Barnea-Goraly et al., 2004; Herbert et al., 2004; Keller, Kana, & Just, 2007). Functional connectivity is defined as the correlation of activation between different brain areas; this measure, though

successful in identifying impaired networks, does not shed light on how the brain areas interact with each other. However, effective connectivity, defined as the influence one system exerts over another with respect to a given experimental context (Büchel & Friston, 2000), allows the estimation of the parameters that control brain activity over time such that activity in one area can be expressed in terms of activity in other brain region(s). With the advent of neuroimaging, studies on emotional and social processing have been able to identify the brain regions that support such tasks. The main brain regions that have been identified as parts of the social-emotional brain include (1) perceptual areas: occipital cortex, fusiform gyrus, and superior temporal sulcus (STS); and (2) emotional processing areas: amygdala, dorsomedial prefrontal cortex (DMPFC), and lateral prefrontal cortex. Each area plays a crucial role in the execution of social-emotional processing. For instance, the amygdala plays a central role in empathy and Theory of Mind understanding. Studies suggest that, in response to increased emotional salience, the amygdala boosts activation towards representational sites of emotional stimuli or events, causing attentional amplification (Phelps & LeDoux, 2005). DMPFC carries emotional information from subcortical limbic regions (including the amygdala) to executive centers in the prefrontal cortex and is involved in top-down regulation of information processing. DMPFC is active in Theory of Mind and state attribution tasks (e.g., Vogeley et al., 2001), during which people with autism consistently demonstrate decreased activation (Happé, 2003). In autism, DMPFC shows minicolumn regulation abnormalities (Casanova, 2004), which Courchesne and Pierce (2005) interpret as local overconnectivity. Lateral prefrontal cortex is a part of the mirror neuron system that is involved in higherorder control processes of cognition and behavior (Dove, Manley, Epstein, & Owen, 2008). It is involved in intention recognition, so an improper functioning will lead to a misunderstanding of others’ intentions and emotions. It has been suggested that dorsolateral PFC (dLPFC) recodes information from posterior association areas into a more abstract form, while ventrolateral PFC (vLPFC) subserves the expression of first-order executive processes (i.e., active selection, comparison, judgment of stimuli). The evidence for perceptual deficits in autism is mixed. Some studies of social-emotional processing have found perceptual impairments in autism when compared to controls (Dalton et al., 2005; Pierce, Müller, Ambrose, Allen, & Courchesne, 2001; Schultz et al., 2003). For instance, Dalton and colleagues (2005) showed that autism and control participants scan and process images of faces differently, potentially leading to the hypoactivation of the fusiform gyrus commonly reported in studies of face processing in autism. Differences in activation between controls and participants with autism during social-emotional processing have also been found in less posterior areas, such as the amygdala and anterior DMPFC (Baron-Cohen et al., 1999; Critchley et al., 2000; Pierce, Haist, Sedaghat, & Courchesne, 2004; Pierce et al., 2001; Schultz et al., 2003; Wang, Lee, Sigman, & Dapretto, 2007).