Evolution of Verb Meanings in Children and L2 Adult Learners Through Reorganization of an Entire Semantic Domain: The Case of Chinese Carry/Hold Verbs

Abstract

The meaning of a word is not acquired in isolation from other words. This article investigates how first-language (L1) and adult second-language (L2) learners of Chinese learn the meanings of verbs belonging to the same semantic domain, focusing on the semantic domain of “carrying/holding” in Chinese. Results revealed that the verb use of L2 adults is heavily influenced by their lexical knowledge of L1 and that their development of word meanings stops before they fully reach the adult native speakers' word meaning. L1 children in contrast tend to depend on perceptually visible features of actions at the initial stage of lexical acquisition and then gradually learn how their L1 categorizes the actions by verbs. We argue that L2 learners need to attain meta-knowledge about the mapping of the entire configuration of the corresponding lexical domain between L1 and L2 and discuss how reading inside and outside of the classroom could foster this process.

ACKNOWLEDGMENTS

This research was supported by Ministry of Education grant-in-aid for Scientific Research (#22243043) to Imai, and Fellowships from the Japan Society of the Promotion of Science (JSPS) and Grant-in-Aid for JSPS Fellows (#232913) awarded to Saji. We are deeply indebted to Henrik Saalbach, Hua Shu, Yuping Zhang, Lianjing Li, and Yuji Kajita for their help with the data collection and discussion. We also thank Jun Shigematsu and Sangwook Lee for their advice and suggestions concerning the linguistic analyses of Chinese and Korean, and Hiroki Okada for writing the program to conduct the experiment.

Notes

¹Throughout the article, to differentiate verbs and the actions denoted by the verbs, we italicized the verbs whereas the referent action for the verb was put in the double quotation marks.

²However, in Saji et al. (2011), the matrices were created only from the carrying videos to simplify the analyses on the ground that the two matrices created for the carrying videos and holding videos were very highly correlated. Here, we decided to create the matrices that aggregated the responses from the carrying and holding videos to increase data points.

³We employed the two dimension solutions for all groups, as the stress value can be considered to be acceptable (.04, .03, .06, .08 for 3-, 5-, 7-year-old children and adults; .02 and. 01 for Japanese and Korean learners; .04 and .01 for native speakers of Japanese and Korean, respectively) for all.

⁴These correlation values are slightly different from what have been reported in Saji et al. (2011) because the responses for the carrying and holding videos were aggregated for the carrying (moving) and holding (nonmoving) actions here, whereas Saji et al. used only the responses from the carrying video.

⁵Some readers may wonder why the produced verbs by Korean and Japanese learners were very similar in Table 3, though they were different in Figure 5A and 5B. This is because the results of MDS reflect the entire response matrices, whereas Table 3 shows a single verb that was produced most frequently.