In this study, the purpose is to use functional MRI scans to investigate which metaphor model, i.e. the conceptual mapping view or the attributive categorization view, better explicates the activation of areas in our brain. In this study, ten right-handed male subjects are presented with 60 sets of metaphorical sentences at a discourse level, in which two conditions are designed: congruent and incongruent. The former indicates that the terminal metaphorical sentence shares the same conceptual mapping with the preceding metaphorical context (i.e. ARGUMENT IS WAR for both the target and context sentences); the latter means the terminal metaphorical sentence shares different mappings with the preceding metaphorical context (i.e. ARGUMENT IS WAR for the target sentence vs. ARGUMENT IS BUILDING for the context). Subjects are instructed to read each sentences carefully and judge whether the terminal metaphorical sentence is acceptable or not within the whole given context. We predict that for the incongruent condition, more effort will be needed to elicit more effort in achieving global coherence (St. George et al. 1999, Robertson et al. 2000) if, in fact, the change in source domains affects processing. In this case, the incongruent condition will recruit larger activation in the right hemisphere than the congruent one, and would support the conceptual mapping model. Otherwise, no significant involvement in the right hemisphere (the null hypothesis) leaves open the possibility that the attributive categorization view is correct. In brief, this neuro-imaging study will detail the metaphor processing locations and will contribute to the understanding of cognitive function in the right hemisphere (Shu-Ping Gong and Kathleen Ahrens)