Abstract

Metaphors have been an object of fascination and a matter of debate since ancient times. What has attracted researchers is how metaphors are so seamlessly understood when their literal meaning differs from what they convey metaphorically. Some scholars have proposed that listeners attain the metaphorical content serially, where the literal interpretation is initially derived, combined with pragmatic information, and then the metaphorical content is determined. Other scholars, however, have argued that the efficiency with which metaphors are understood does not allow for the literal meaning to be derived first and then rejected in favour of the metaphorical content. Rather, they contend that most conventional metaphors are directly retrieved from semantic memory without the need for any inferential work. This thesis presents three manuscripts that investigated two-word metaphors such as broken heart and sharp tongue. The first manuscript reported a norming study to serve as an open source of materials required to run experiments such as those in the current thesis. The second manuscript examined whether the literal meaning of conventional metaphors was available, and could be recovered, immediately after the metaphorical content had been attained. In a maze task, participants read sentences word by word in a self-paced manner and then choose which of two words correctly continues the sentence, where the distractor words were either related or unrelated to the metaphorical content of the sentence. The results of this study yielded a significant awakening effect, whereby longer response times and lower accuracy rates were obtained in trials in which the literal meaning was cued immediately after the metaphor had been processed. This pattern of results suggests that the literal meaning was awakened during sentence processing. The third manuscript examined whether the awakening effect could be found further away from the metaphorical expression. The results of this study also yielded a significant awakening effect. However, it was weaker when compared to the original maze. Lastly, in Appendices A and B, the effects of familiarity and aptness on the awakening effect were analyzed. The results of these analyses indicated that, when the literal meaning is cued immediately after the metaphorical expression has been processed, familiarity and aptness do not have an overall effect. However, further downstream, the awakening effect is indeed modulated by familiarity and aptness. Altogether, the results from the series of studies presented in the current thesis provide compelling evidence in support of the literal-first approach.