From the age of 18 months on, toddlers learn new words at the phenomenal rate of over 9 a day| To support such successful learning, children must have considerable knowledge of the nature of the meanings words encode. This research will seek a precise characterization of that knowledge and of its roots in early infancy. The research will focus on one aspect of word meaning, that of quantification. Language makes use of many different types of quantificational devices, including quantifiers ("a, another, some, more...."), the numerals, and count/mass syntax. Knowledge of quantification would help even the learning of nouns, for nouns carry quantificational commitments in their meanings. Children use quantifiers, apparently correctly, very early in language production. Recent research has shown, however, that children do not know what the count words mean for over a year after they have been counting. This raises the possibility that they do not know what the other quantifiers mean either. The studies will use new techniques to assess what even babies know about number. Further, the techniques will allow the assessment of patterns of generalization to new contexts of various types of newly-taught words, supported by different sorts of quantifiers. A longitudinal study will assess the process by which two-year-olds work out the meanings of number words. The outcome of these studies will be an understanding of the conceptual underpinnings normal young children bring to language learning, and the conceptual change they must undergo to grasp the meaning of number words like "two." Research on normal cognitive development and normal language development is important scientifically, for it bears on what makes humans unique among animals. But it is also essential for understanding the many ways language acquisition goes awry in various abnormal populations. Such understanding must precede interventions to help children overcome specific cognitive deficits.
