Only a few studies have produced evidence of similar problems being found in subjects with Asperger Syndrome. Baron-Cohen et al. reported no impairment in identifying basic emotions (e.g. happy, sad) in their group of high-functioning adults with autism and Asperger Syndrome compared to controls. However, the authors reported a significant deficit in these individuals in identifying more complex mental states (e.g. interest, surprise, embarrassment). Similarly Grossman et al. investigating emotion processing in AS found no impairment in a basic emotion recognition task. They found, however, an Asperger specific impairment compared to normal controls in face recognition when a mismatching emotional label was paired with the facial stimulus. The authors interpreted their finding that qualatative differences exist in how individuals with AS process facial emotions and suggested that they use compensatory strategies, primarily verbal mediation in emotion recognition in laboratory tests. In simple recognition of facial emotions, they can hide their problems by using these strategies. However, when faced with more difficult conditions, as in the above taks employing mismatching words, these strategies are not always satisfactory.

The compensatory process of verbal mediation and also possible other strategies could help explain the pervasive oddness and rigidity in social life that characterizes individuals with Asperger Syndrome. There are many clinical and anecdotal reports suggesting that individuals with Asperger Syndrome do not possess an intuitive understanding of other people's affective states but rather use their good analytical or verbal abilities to find out what a person might be feeling. These compensatory tactics may work satisfactorily in experimental situations but are largely not very successful in complex social exchanges in everyday life. As pointed out by Grossman et al., 'This cognitive strategy might go far to explain the lack of spontaneity and seemingly "rehearsed" quality characteristic of AS social interactions. Even most highly functioning adults with autism and AS appear to follow overlearned "scripts" to guide them through their social interactions, which are often used in a rather rigid fashion that is seen by others as unnatural or robotic'.

Abnormalities in brain activation in persons with Asperger Syndrome and very high-functioning autism during tasks involving the perception of facial expressions and during TOM tasks have also been reported in two functional MRI studies. Based on this research evidence Baron-Cohen and colleagues suggested that the amygdala is one of several neural regions which are not normal in autism and put forward an 'amygdala theory of autism'. These imaging results further support the suggestion that unusual cognitive strategies are employed in performing these taskes.

Two types of cognitive processing seem to be involved in ther recognition of emotion, firstly, analytical processing, which involves the left hemisphere and secondly, holistic processing, which depends on the right hemisphere and subcortical processes including the limbic system. It seems that individuals with Asperger Syndrome have difficulties in social interactions not because of a major deficit in perception comprehension of facial expressions but becuase they use analytic and verbally mediated strategies rather than a holistic and intuitive processing found in normal and also clinical subjects.