Modeling the Development of Infant Categorization

Research on infant categorization has made remarkable progress since the first studies were reported in the late 1970s (e.g., Cohen & Caputo, 1978; Cohen & Strauss, 1977; Strauss, 1979). This progress is evident in a recent volume on early categorization and concept acquisition (Rakison & Oakes, 2003), the first half of which is devoted entirely to theory and research on infant categorization. Even in the early days of such research an important distinction was made between demonstration‐oriented studies and process‐oriented studies (Cohen & Younger, 1983; Younger & Cohen, 1985). Demonstration studies simply presented infants with established category items (by adult standards) such as pictures of stuffed animals, faces, dogs, cats, animals, or vehicles and examined whether infants would generalize their responding to novel members of the same category. Process studies, on the other hand, presented infants with novel categories and manipulated feature values of category items to examine the mechanisms underlying infant category acquisition.