University of Minnesota
Driven to Discover


Learning Categories and Classifiers

Classifiers are words used in languages such as Mandarin Chinese when counting and referring to objects. Hmong, Japanese and Mandarin are classifier languages and examples with classifiers in these languages are given below:

Hmong: ib daim nplooj ‘one/a-CLASSIFIER-leaf’
Japanese: happa iti mai ‘leaf -one-CLASSIFIER’
Mandarin: yi pian yezi ‘one-CLASSIFER-leaf’

We have collected evidence on how speakers of Hmong, Mandarin, and Japanese use classifiers with the most commonly used nouns for solid objects (Sera, Soh, Fuller, Stevens, Batteen, Kuo, & Davis (2009). We have also shown that adult speakers of Mandarin Chinese rely more heavily on shape than English speakers when classifying solid objects, and that this difference is likely caused by language (Kuo & Sera, 2009). In future work, we will be using a two-pronged strategy to further understand the relations between developing language and cognition:

  1. We have begun to apply computational (connectionist) models of learning to the task facing children who are learning concepts as they learn classifiers (Sera & Stevens, 2009). These simulations follow the simulations conducted by Rogers and McClelland (2004) on category learning with the exception that our simulations incorporate classifiers into the model. The results from these simulations should offer a better understanding of the dynamics between developing words and concepts and they will also provide a principled framework from which to ask additional empirical questions. We plan to expand these simulations in future work.
  2. We are attempting to link developing knowledge of the classifiers ge, zhi, and tiao to the differences in categorization between English- and Mandarin-speaking adults found by Kuo and Sera (2009).  In this ongoing work, we are testing Mandarin speakers in both a classifier comprehension task and a categorization task, and comparing their results in the categorization task to age-matched English-speakers.  Future work will extend this strategy to speakers of other classifiers languages.

Collaborators: Hooi Ling Soh, Kaitlin Johnson and James Stevens.

Child participating in classifiers testing.


© 2009 Regents of the University of Minnesota. All rights reserved.
The University of Minnesota is an equal opportunity educator and employer
Last modified on November 20, 2008