Processing Cantonese classifiers by Mandarin speakers: an eye-tracking study
Other conference paper

Full Text

Other information
AbstractChinese has a rich set of nominal classifiers which collocate with nouns in sentences. The classifier-noun collocation is not random. Previous studies have shown that native speakers of Chinese predict the upcoming noun based on the classifier they hear or read. In Cantonese, classifiers can quantify a noun without any numerals and demonstratives. The ‘bare classifier phrase’ ([CL-N], e.g. zi1 bat1 ‘CL pen’) has a definite reading when appearing in a pre-verbal
subject position, whereas Mandarin does not typically allow [CL-N] in the same position (Li & Bisang 2012). Previous studies have shown that second language (L2) learners demonstrate predictive understanding in real-time processing as native speakers do in some areas, albeit to different extents (e.g. Lau & Grüter 2015). This study investigates whether Mandarin-speaking learners of Cantonese are able to predict the upcoming noun based on the classifier they hear through an eye-tracking experiment. Adopting the Visual World Paradigm (e.g., Altmann & Kamide, 1999), we asked native Mandarin speakers with relatively low L2 Cantonese proficiency (n = 17) to look at four pictures depicting different daily objects while listening to Cantonese sentences with [CL-N] in the subject position. Their eye-gaze patterns were recorded and compared with those of native speakers of Hong Kong Cantonese (n = 18). Participants also completed two general cognitive ability tasks (to measure executive function and working memory) and a postexperiment classifier-noun collocation test. Results show that the L2 speakers were able to fixate onto the target noun 300 milliseconds after the offset of the classifier, regardless whether the classifier-noun collocation resembles their L1 Mandarin or not. Although L2 speakers performed poorer than the native speakers in the post-experiment classifier-noun collocation test, their performance in the eye-tracking experiments was not statistically different from the native speakers. This finding supports the proposal that native and nonnative processing are not fundamentally different when the task involves establishing and checking local, rather than long-distance dependency, in this case, within the determiner phrase (DP). Nevertheless, the delayed anticipatory looks at the target nouns found in both
native and non-native speakers give rise to questions with regard to the primary function of Chinese classifiers in real-time processing and point out new directions for future research.
All Author(s) ListLEE Ka Chun, KANG Xin, MAI Ziyin
Name of ConferenceThe Linguistic Society of Hong Kong Annual Research Forum 2016
Start Date of Conference03/12/2016
End Date of Conference03/12/2016
Place of ConferenceHong Kong
Country/Region of ConferenceHong Kong
LanguagesEnglish-United Kingdom

Last updated on 2018-22-01 at 12:48