中華心理學刊
民96 ， 49 卷， 3 期， 301-314

David Allbritton（Department of Psychology, DePaul University）

Abstract

Previous research on the effects of metaphor on text comprehension provided evidence that conceptual  metaphors can connect parts of a text representation, as evidenced by priming in word and sentence recognition
tests. This study extended the previous findings with a different priming methodology, and investigated whether forming connections depends on the conventionality (familiarity) of the metaphors used. Although the effects of
metaphorical consistency were not robust (significant priming effects were evident only after post-hoc filtering of the data to remove subjects who did not attend to the task), both novel and conventional metaphors were found to produce priming for metaphor-related words and sentences, suggesting that conventionality was not necessary for the metaphors to make connections. When brief (250 ms) prime presentations were combined with short (2000 ms) response deadlines in a word recognition task, however, conventional metaphors produced
priming but novel metaphors did not. The relationship of the results to the Class Inclusion and Conceptual Metaphor theories of metaphor is discussed. 

Keywords: metaphor, conceptual metaphor, text comprehension, schemas