A central point that has emerged from the chapters on null subjects is that although children universally omit subjects, there is also an interesting range of cross-linguistic variation in the young child's knowledge and use of null and overt pronouns. Pierce (this volume, chapter 14) notes, for example, that subject pronouns in early French and early English differ in distribution and frequency. Valian (this volume, chapter 11) observes that although American and Italian children both use null subjects, Italian children have a much higher frequency of null subject use than English-speaking children at a comparable developmental stage. Lillo-Martin (this volume, chapter 13) points out that although both English- and Chinese-speaking children have null subjects, only Chinese-speaking children drop objects, as is permissible in the adult language (cf. Wang, Lillo-Martin, Best, & Levitt, 1992). Finally, Rizzi (this volume, chapter 10) claims, based on claims in Valian (1991) and Roeper and Weissenborn (1990), that null subjects in early English are restricted to root contexts, whereas in null subject languages such as Italian, they are not so restricted. What these examples show is that child languages, like adult languages, show a complex interaction of universal and language-specific, that is, input-determined, properties.