Social cognition broadly refers to the way people make sense of others and themselves (Fiske & Taylor, 1991), the way people construct repre-sentations of the relation between self and other, and the way people flexibly use such representations to guide social behavior (Adolphs, Sears, & Piven, 2001). Human social interactions require the ability to relate self with others, as well as the ability to distinguish self from others. However, a key question is at what stage in development do we begin to differentiate others from ourselves, and what role does this distinction play in the development of critical social-cognitive functions, such as Theory of Mind (ToM), action

understanding, embodiment, social evaluation, and empathizing? Indeed, these social-cognitive functions appear to be interrelated, as ToM refers to the ability to understand that mental states can cause, explain, and predict others’ behavior (Premack & Woodruff, 1978), and action understanding, and embodiment may be basic to mental state attribution. Although embodiment and empathizing are mechanisms that are thought to allow us to see others as “like me” (Meltzoff, 2007), it is possible that the development of the concept of “me” facilitates the development of embodiment and empathizing, and that an atypical representation of “me” inhibits these processes, thus inhibiting ToM. Therefore, a second question pertains to learning the nature of typical functions from the dysfunctions in disabled populations, such as neurodevelopmental disorders. For instance, autism spectrum disorders (ASD) have been found to be characterized by extreme withdrawal to oneself, complete self-focus, or extreme egocentrism (Kanner, 1943; Asperger, 1944; Frith & de Vignemont, 2005). Examining the neural correlates of self in individuals with ASD may help us gain insight into important social-cognitive functions like ToM, social evaluation, and empathy. To this end, we will examine the integrity of three neural systems (cortical midline structures, the mirror neuron system, and the Theory of Mind network) thought to underlie self-related thinking and related aspects of social cognition and social functioning. Then, we will examine how these neural systems are possibly altered in individuals with ASD. For self-understanding, we will examine cortical midline structures (CMS), a set of regions in the midline of the cortex arching around the corpus callosum. The CMS (constituting the medial prefrontal cortex, the anterior, middle and posterior cingulate cortices, and the precuneus) are thought to functionally integrate self-related thought and planning (see Northoff et al., 2006). The CMS seem to attenuate their activity when participants engage in explicit, attentiondemanding, and goal-directed cognitive tasks. These structures are also called the default mode network of the brain, which may be the psychological baseline of brain functioning (Raichle, 2001; Gusnard & Raichle, 2001; Fox & Raichle, 2007). The activity of this network has been linked to task-unrelated mental processes (Giambra, 1995). Such thoughts are usually autobiographical in nature, which would naturally have strong associations with the self. fMRI and positronemission tomography (PET) studies have implicated the CMS in self-referential processing compared to non-self-referential tasks (see Northoff et al., 2006, for a meta-analysis of brain activation in several studies on self-processing). In addition to playing an active role in tasks that focus on self, most of the CMS regions have been found to be active during resting state, when people tend to reflect on themselves and their actions (Raichle et al., 2001; Gusnard & Raichle, 2001; Fox & Raichle, 2007; Raichle & Snyder, 2007). In autism, resting state fMRI studies have revealed altered neural response during self-referential thinking (Cherkassky, Kana, Keller, & Just, 2006; Kennedy, Redcay, & Courchesne, 2006), which, in addition to the behavioral evidence of abnormalities in the development of self-concept, provide a compelling case to study self in persons with ASD.