Prototype abstraction and distinctive feature learning: an application to learning Chinese characters created by Noriyuki Matsuda and Donald Robbins
Material type:
- text
- unmediated
- volume
- 0144-3410
- LB1051 JOU
Item type | Current library | Call number | Vol info | Status | Notes | Date due | Barcode | |
---|---|---|---|---|---|---|---|---|
![]() |
Main Library - Special Collections | LB1051 JOU (Browse shelf(Opens below)) | Vol. 69, no.1 (pages15-23) | Not for loan | For in house use only |
Chinese characters with related meanings often have a common symbol included among their components. In an exemplar-prototype and distinctive feature classification system, the multiple-component characters (the exemplars) can be said to have a common component (the prototype) distinctive to the shared concept. A study was conducted with 60 undergraduates who were administered recognition tests with new and old exemplars and prototypes. Results show that the traditional language learning technique of paired-associate training with exemplars of Chinese characters and specific English translations led to the poorest performance of the 3 methods tested. Learning either the English concept for Chinese exemplars (exemplar-category training) or the English concept for only the distinctive feature (prototype training) produced greatly superior performance. Instructions regarding the nature of Chinese characters had no effect on performance.
There are no comments on this title.