<P> Research on theory of mind in autism led to the view that mentalizing abilities are subserved by dedicated mechanisms that can - in some cases - be impaired while general cognitive function remains largely intact . </P> <P> Neuroimaging research has supported this view, demonstrating specific brain regions consistently engaged during theory of mind tasks . PET research on theory of mind, using verbal and pictorial story comprehension tasks, has identified a set of brain regions including the medial prefrontal cortex (mPFC), and area around posterior superior temporal sulcus (pSTS), and sometimes precuneus and amygdala / temporopolar cortex . Subsequently, research on the neural basis of theory of mind has diversified, with separate lines of research focused on the understanding of beliefs, intentions, and more complex properties of minds such as psychological traits . </P> <P> Studies from Rebecca Saxe's lab at MIT, using a false - belief versus false - photograph task contrast aimed at isolating the mentalizing component of the false - belief task, have very consistently found activation in mPFC, precuneus, and temporo - parietal junction (TPJ), right - lateralized . In particular, it has been proposed that the right TPJ (rTPJ) is selectively involved in representing the beliefs of others . However, some debate exists, as some scientists have noted that the same rTPJ region has been consistently activated during spatial reorienting of visual attention; Jean Decety from the University of Chicago and Jason Mitchell from Harvard have thus proposed that the rTPJ subserves a more general function involved in both false - belief understanding and attentional reorienting, rather than a mechanism specialized for social cognition . However, it is possible that the observation of overlapping regions for representing beliefs and attentional reorienting may simply be due to adjacent, but distinct, neuronal populations that code for each . The resolution of typical fMRI studies may not be good enough to show that distinct / adjacent neuronal populations code for each of these processes . In a study following Decety and Mitchell, Saxe and colleagues used higher - resolution fMRI and showed that the peak of activation for attentional reorienting is approximately 6 - 10mm above the peak for representing beliefs . Further corroborating that differing populations of neurons may code for each process, they found no similarity in the patterning of fMRI response across space . </P> <P> Functional imaging has also been used to study the detection of mental state information in Heider - Simmel-esque animations of moving geometric shapes, which typical humans automatically perceive as social interactions laden with intention and emotion . Three studies found remarkably similar patterns of activation during the perception of such animations versus a random or deterministic motion control: mPFC, pSTS, fusiform face area (FFA), and amygdala were selectively engaged during the Theory of Mind condition . Another study presented subjects with an animation of two dots moving with a parameterized degree of intentionality (quantifying the extent to which the dots chased each other), and found that pSTS activation correlated with this parameter . </P>

When children develop theory-of-mind (tom) they can recognize that