Further Comments on the Attribute Availability Hypothesis of Children's Category Learning

Kemler Nelson raises several important issues regarding children's category learning in her commentary on Ward, Vela, and Hass. The points of agreement in our positions concern the extreme complexity of the phenomenon. The points of disagreement that are addressed in this reply concern Kemler Nelson's assertions that (a) children are, in general, holistic processors, (b) a holistic processing model provides a better account of children's category learning than the attribute availability model, and (c) that Ward et al. have overestimated the extent of analytic processing and underestimated the extent of holistic processing by young children. All of these points are challenged in the present paper.