Neural Correlates Of Theory-of-mind Reasoning: An Event-related Potential Study

Theory-of-mind (ToM) which plays an important role in human social interactions is an ability to understand how other's behaviors are motivated by their internal mental states, such as beliefs, desires, intentions, and emotions. It is an area of much concern among developmental psychologists. The development of neuroimaging technology provides the tools and timing to investigate neural substrates of theory-of-mind reasoning, which became one of the research focuses of cognitive neuroscience.Because of the relatively low temporal resolution of PET and fMRI, the time-course of cerebral dynamic processes can't be precisely described and explained. As a result, previous studies emphasized particularly the cerebral localization of theory-of-mind reasoning, but neglected the analysis of the time-course in the dynamic processes. Based on the literature review of former studies, the dissertation is to examine the brain mechanism of the dynamic processes of theory-of-mind reasoning. By pointing out the defects of previous studies in this area, this dissertation exerted the advantage of high temporal resolution of ERP, recorded the EEG data from 32 scalp electrodes by ERP system (Neuroscan Inc.). Through employing the normal undergraduates and graduates as participants, the present dissertation analyzed the behavioral performance (accuracy and reaction times), ERP components and topography off-line to reveal the time course of the theory-of-mind reasoning.The present dissertation consists of three ERP studies. The first one is designed to examine the dynamic dissociation of the representation of another person per se and the representation of that other person's mental states, which are included in the theory-of -mind reasoning. The latter two studies are to explore the time-course of the process of some mental states composing the theory-of-mind, such as desire, belief. Study one employed the cartoon comprehension task to examine the dynamic process of reasoning of theory-of-mind. In order to further decompose the sub-components of theory-of -mind, study two used the revised deceptive-appearance task to investigate the differences of the neural basis between the belief vs. desire reasoning. Study three examined the neural correlates of understanding self and other's mind. The results provide the ERP indexes of dynamic process of theory-of-mind reasoning. The three conclusions have been drawn as follows:1. To investigate the dynamic dissociation of theory-of-mind reasoning and person perception, ERPs were recorded while the subjects were watching three kinds of cartoons. The amplitudes of P1 for scene pictures were significantly lower than person perception cartoons and ToMC. There were no differences in the amplitudes of P1 between the two latter conditions. The differences between ToMC and no-ToMC started from 1000ms post-stimulus to 1300ms. ERP waveforms showed effects of dynamic dissociation in the LPC range: the LPC was more positive in ToMC condition relative to no-ToMC condition; while no-ToMC condition elicited a more positive ERP deflection than no-humc. The present study provided electrophysiological evidences of dynamic dissociation of the person perception and reasoning the mental states in theory-of-mind. While a representation of a person per se is a likely prerequisite for ToM, achieving a representation of others'mental states is the core responsibility of a ToM. These findings revealed the dynamic process to distinguish between animate and inanimate entities, and suggested that theory-of-mind reasoning possibly begin with the person perception or the scene decoding, and then achieve a representation of others'mental states.2. Late negative component (LNC) was generated in centro-frontal areas during the other's desire vs. belief reasoning. Starting at 350ms post-stimulus, the mean amplitudes of LNC for the belief condition were more negativity than LNC for the desire condition. These differences were maintained through 1000ms. It is possible that the dissociations reflect the discrepant processes associated with other's desire vs. belief reasoning. Further, it took participants much more time to reason about the belief questions. The data from the reaction time suggests that it is easier to reason about other's desire than belief. It is possible that the reasoning of belief and desire belongs to different levels in theory-of-mind, and may rely on dissociable neural mechanism.3. LNC was elicited during understanding representational change questions (self mind) vs. false belief questions (other's mind), which amplitudes were significantly different for two conditions. Both the time course and the spatial distribution of the dissociations are noteworthy. Results indicated that the mean amplitudes of LNC elicited by the two conditions differed at the following two time windows: 350-450ms and 650-1050ms. False belief reasoning (other's mind) was associated with a stronger negativity than the representational change question. The dissociation in 350-450ms may indicate that the subjects can distinguish between the one's own mind and other's mind. The difference of the LNC in 650-1050ms indexed that understanding one's own mind and other's mind may rely on dissociable neural mechanisms. With respect to spatial distribution, the LNC in the middle scalp are more negative than that in the post area.