Are there principles that apply only to the acquisition of words? A reply to Waxman and Booth.

Young children are strikingly good at learning the meanings of words, even under conditions of limited exposure. In Markson and Bloom (1997), we suggest that this `fast mapping' ability is not special to word learning. Children who heard a novel object described, in passing, as `a koba' remembered which object the word referred to when tested a month later. However, children who heard a novel object described, in passing, as `the one that my uncle gave me' were equally good at remembering which object the fact referred to. Waxman and Booth (2000) replicate this ®nding, but they also found a difference between the word and the fact. They found that children extended the word to novel objects from the same category, but did not show this pattern of extension for the new fact. We ®nd this result unsurprising. Numerous studies have shown that children view count nouns as referring to categories (e.g. Brown, 1957; Macnamara, 1972; Waxman, 1999). And so when asked `Is this one a koba?' and `Are there any other ones that are kobas?', children will choose other objects that fall into the same category as the object originally called `a koba'. But there is no reason to expect them to do so when asked `Is this one my uncle gave me?' and `Can you show me one that my uncle gave me?' The fact that a given object is a gift from the experimenter's uncle is information about a speci®c individual, and it would be perverse for a child (or an adult, for that matter) to extend it to other objects from the same category. If one is going to compare words and facts, the appropriate match for such a fact would not be a common noun ± it would be a proper name, which is an expression that also picks out a speci®c individual. P. Bloom, L. Markson / Cognition 78 (2001) 89±90 89