EEG evidence of face-specific visual self-representation

Cognitive science has regarded an individual's face as a form of representative stimuli to engage self-representation. The domain-generality of self-representation has been assumed in several reports, but was recently refuted in a functional magnetic resonance imaging study (Sugiura et al., 2008). The general validity of this study's criticism should be tested by other measures to compensate for the limitation of the time resolution of the blood-oxygen-level-dependent (BOLD) signal. In this article, we report an EEG study on the domain-generality of visual self-representation. Domain-general self-representation was operationally defined as the self-relevance common to one's own Face and Cup; three levels of familiarity, Self, Familiar, and Unfamiliar, were prepared for each. There was another condition, Visual Field, that manipulated visual hemifield during stimulus presentation, but it was collapsed because it produced no interaction with stimulus familiarity. Our results confirmed comparable phase resetting in both domains in response to familiarity manipulation, which occurred within the medial frontal area during 270-390 ms poststimulus and in the theta band. However, self-specific dissociation was observed only for Face. The results here support the conclusion that visual self-representation is domain-specific and that the oscillatory responses observed suggest evidence of face-specific visual self-representation. Results also revealed an inter-trial phase coherency decrease specifically for Self-Face within the right fusiform area during 170-290 ms poststimulus and in the alpha and theta band, suggesting reduced functional demand for Self-Face represented by sharpened networks.