Published online by Cambridge University Press: 26 September 2008
Distinctive feature specification and representation in phonological acquisition are examined in the context of underspecification theory. Subjects were 30 children (aged 3;1 to 5;10) who exhibited systematic differences in their linguistic knowledge of target phonological contrasts. A free classification task was used to tap children's conceptual knowledge of these contrasts, with features of place and manner experimentally manipulated. Three questions were addressed: which features do children use to categorize segmentai information, do the defining features of a category shift as the phonological system advances, and which framework of underspecification theory best accounts for the results? All children categorized segments on the basis of marked nonredundant featural properties, and used only one feature value to define category membership consistent with radical underspecification. Linguistic knowledge and linguistic input both influenced children's category judgements, but to different degrees. The emergence of phonological categories involved increasing feature differentiation as the child's productive phonology advanced.
This research was supported in part by grants from the National Institutes of Health to Indiana University (DC 01694, DC 00076). Jessica Barlow, Daniel Dinnsen, Mary Hughes and Michele Morrisette provided insightful comments and assistance. The following research assistants also participated in various aspects of data collection, analysis, and phonetic transcription: Annette Champion, Steve Chin, Elizabeth Connell Antonnen, Jennifer Huljak, Holly Storkel and Jennifer Taps. Aspects of this paper were presented at the International Conference on Phonological Acquisition, Vancouver, British Columbia, June 1995.