| There is much controversy about the roles of topics and vehicles in the process of metaphor comprehension. Comparison models claimed that metaphor comprehension began with the alignment of the topics. Categorization models maintained that metaphors were comprehended by assigning the topic to an abstract category associated with the vehicle. According to the career of metaphor model, as a vehicle became conventionalized, the comprehension of metaphors would switch from alignment-first processing to abstraction-first processing.The present research was an online experiment designed to test whether there was such a shift in comprehension of Chinese metaphors in the form of "A是B" (Noun A is Noun B). The study followed a 2 (vehicle conventionality: high and low)×2 (similarity: high and low)×3 (priming types: vehicles, topics and pattern masks) within-subject design.It was found that priming by vehicles led to significantly faster comprehension of metaphors with highly conventional vehicles and low similarity, but not of those with highly conventional vehicles and high similarity. The priming effects of vehicles and topics were not significantly different in comprehending metaphors with low conventional vehicles. The results, therefore, provided evidence for a shift from alignment-first processing to abstraction-first processing with the increasing conventionalization of vehicles. However, the shift was not complete, suggesting that other factors (in the present study, similarity) apart from conventionality might be involved in the shift. |