2018 Friday Poster 6393

Friday, November 2, 2018 | Poster Session I, Metcalf Small | 3pm

Deciding the Referent of a New Word: The Acquisition of Classifiers
W. Ma, P. Zhou

Upon hearing a new word, listeners are faced with an inherently complex task of deciding the referent of the word. One useful cue to the meaning of a new noun is the classifier preceding the noun. Although classifiers are not generally a feature of English or other European languages, classifier-like constructions are used with certain nouns in English. Thus, X in “a sheet of X” should refer to a broad, flat piece of object, while Y in “a chunk of Y” most probably refers to a thick, solid object. Classifiers can be exceptionally important for word learning in languages like Chinese where classifiers are obligatory when the number of entities needs to be specified. For example, the phrase “three people” should be expressed as “three classifier people” in Chinese.

This study aims to 1) determine if 3- to 5-year-old Chinese learners can use classifiers to determine the referents of new words, and to 2) document the development of classifiers. The participants were 3-, 4-, and 5-year-old monolingual Chinese learners (n=24/age-group). We used 16 classifiers referring to shape, vehicle function, human-animal distinction, and arrangement, respectively (Table 1).

Using the intermodal preferential looking paradigm, an experiment contained sixteen 6-s tests (one per classifier). On each trial, children saw two novel images side-by-side (e.g., two objects, or two vehicles, or two arrangements, or an animal-human pair). For the accompanying speech stimuli, the onset of the classifier divided a trial into two phases: a pre-target phase and a post-target phase (“Kan4 [look]! Zhe [This] shi [is] yi [one] classifier shenme [an existential indefinite like something]). A significant increase in looking time to the target image across phases indicates that the participant has found the target image. As the existential indefinite shenme is uninformative of the target, children must rely on their knowledge of classifiers to find the target image in this task (Table 2).

Mean increases in looking time at the target image across phases were calculated for each type of classifiers. Results showed that the 3-year-olds reliably used the shape (p=.005) and human-animal distinction classifiers (p=.002) to find the target images. However, their looking time at the target image did not increase across phases on trials of vehicle function classifiers or arrangement classifiers, suggesting that they did not have a robust understanding of these classifiers. The same pattern of results was found in the 4-year-olds. The ability of using vehicle function classifiers was observed only in the 5-year-olds (p=.012). However, even for the 5-year- olds, the increase in looking time at the target image across phases was only marginally significant on the arrangement classifier trials (p=.028; an adjusted significance cutoff level of .0125 was used throughout this study).

Thus, Chinese children can determine the referents of new words solely depending on classifiers as early as age 3. Classifiers therefore can be an important support for child word learning. Furthermore, the development of classifiers continues even beyond age 5 in a classifier language.