Embodiment in Action: An Empirical Study on the Processing of Chinese Action Metaphors and Idioms

Abstract

This research investigates how the metaphor and idiom are understood and conceptualized by native Chinese with the empirical approach. Based on a large body of previous research, it is well accepted that the sensorimotor system plays an essential role in language processing (e.g., Glenberg & Kaschak 2002, Zwaan & Pecher 2012), concrete language processing in particular. However, it is still under debate whether the understanding of abstract concepts can be fully explained by embodiment theories. To address the important issue of abstract concept understanding, the current study focuses on the processing mechanism of action metaphors and idioms. Action metaphors and idioms are commonly known as verb phrases that consist of a concrete verb but express abstract meanings (i.e., action metaphor: qiān yīnyuán牵姻缘, to make a match between two people; action idiom: diū fànwǎn丟飯碗, to lose one’s job; tuō hòutuǐ拖後腿, to hinder somebody), which makes it a bond of concrete and abstract languages.

Study 1 adopts the behavioral approach using an adapted double-phrase lexical decision task. Through the observation of participants’ reactions to concrete vs. metaphorical vs. idiomatic action verb phrases (VPs) when they were primed with concrete actions using either the same or different effector(s), this study tries to figure out whether the processing of action metaphors and idioms would be affected (either facilitated or interfered) by the pre-activated actions of the same effector. After a series of normative tests, 144 concrete context phrases (i.e., 48 literal hand-related VPs, 48 foot-related VPs, and 48 nouns) and 144 target verb phrases (i.e., 48 concrete VPs, 48 action metaphors, and 48 action idioms) were selected. The participants were presented with a context phrase which was followed by a target phrase, and they were required to make a prompt and accurate judgment on whether the two phrases in each trial were both true words. Results of Study 1 did not show significant differences in the reaction times across conditions but highlighted the role of familiarity in simple tasks. This study did not show strong supporting or opposing evidence to the embodied view of abstract language processing.

Study 2 was designed on the basis of Study 1. A lexical decision task may only require surface activation of the meaning of each word, which may not be enough for the sensorimotor system to take effect; it is also possible that the influence from the recruitment of the sensorimotor system simply did not manifest on the reaction times of a lexical decision task. Therefore, in Study 2, a double-phrase semantic judgment task was implemented to make sure the full meaning of each word was activated in the experiment. In this study, participants made judgments on whether the two sequentially presented phrases recruited the same body part (i.e., hand/arm, foot/leg, or mouth). When all were primed with literal hand-related verb phrases across conditions, participants’ reaction times to metaphorical vs. idiomatic vs. mental vs. literal foot/leg-related action VPs were recorded and compared. Results showed that mental VPs received the fastest responses whereas literal foot/leg-related VPs required the longest reaction times. Metaphorical and idiomatic VPs did not show significant differences in reaction times. The results supported the embodied view of language processing including abstract languages (i.e., action metaphors and idioms). The significantly slower response times to action metaphors and idioms vs. mental VPs suggest that the hand-related motor system is activated during the processing of action metaphors and idioms so that it takes longer to deny that a literal hand-related action was mentioned in those VPs; otherwise, the reaction times of action metaphors and idioms should be similar to those of mental words. As for the longest reaction times to foot/leg-related VPs, it indicates that the activation of a different motor area might require extra time as compared to the activation of the same motor area twice. This result also provides supportive evidence for the embodiment theories of language processing.

Therefore, it is concluded that the embodied view of language understanding can explain the processing mechanism of not only concrete VPs but also action metaphors and idioms. However, how or to which degree the sensorimotor system plays the role in this process may vary; different experimental tasks, even a slight change, may also affect the result. The findings of this research may shed new light on the theoretical development of the embodiment of abstract concepts.

In brief, this thesis have three main contributions to the field of abstract concepts processing: i) this thesis is the first pilot study that looks into the Chinese action-based abstract language processing from the angle of action metaphors and idioms processing, since no study to date has used Chinese figurative stimuli to investigate this unsettled issue of abstraction under the embodiment framework; ii) it demonstrates that task difference plays a significant role in the process of abstract reasoning and suggests that more attention should be paid to the experiment design and task selection in this vein of study; iii) this thesis also underlies the uniqueness of Chinese language since the research findings of the current study slightly divert from the existing literature using European languages as stimuli, which draws more attention to the language typology in the selection of experiment stimuli in future studies.

Date of Award	11 Apr 2024
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Meichun LIU (Supervisor)