Abstract

Several recent papers propose that child and adult grammars differ in their underlying representations of universal quantification, e.g., “every” in English. These proposals attempt to explain children’s nonadult responses, in certain circumstances, in response to sentences that contain the universal quantifier. Blaming children’s nonadult behavior on their grammars is questionable, however, in view of the restrictiveness of the theory of Universal Grammar, which tightly constrains the hypothesis space children can navigate in the course of language development. The restrictiveness of the initial state has led to the Continuity Hypothesis, i.e., the proposal that child language can differ from the local language only in ways in which distinct adult languages can differ from each other. Advocates of the Continuity Hypothesis point out that violations of universal constraints in children’s grammars would have far reaching negative consequences, resulting in myriad ‘errors’ by children both in production and comprehension. Before embracing an analysis that involves such violations, therefore, evidence that children make such errors is needed. In this paper, we draw out the implications of a recent account of children’s nonadult interpretation of universal quantification, by Geurts (2004). We demonstrate that children commit few, if any, any of the ‘errors’ that are predicted by the account, beyond the original nonadult behavior that it was specifically designed to explain. Moreover, the proposed analysis of children’s nonadult grammar is demonstrably inconsistent with their proven semantic competence in interpreting a variety of utterances containing the universal quantifier, alone and in combination with other logical expressions (e.g., disjunction, negative determiner phrases, etc.). These findings buttress previous alternative accounts of the observed differences between children and adults, thereby leaving the Continuity Hypothesis unscathed.