Children do not ignore (null objects): Against deficit accounts of the null object stage in language acquisition

Abstract

Children across a variety of languages omit direct objects at higher rates that adults. It has been argued that these omissions arise from children’s performance or pragmatic limitations. The null object approach holds that children start by allowing a broader set of mechanisms for the recoverability of null objects than those possible in the adult grammar, which becomes more restricted with experience. Comprehension data is considered key evidence for evaluating representational approaches, but the interpretation of previous comprehension results is obscured by methodological issues. This article presents new data contrasting the interpretation of various types of direct objects in negative sentences, including null objects (Johnny is not eating) and anaphoric and negative polarity items (not eating it/not eating anything). English-speaking children aged 4–5 (n = 75) participated in three separate comprehension studies contrasting the interpretation of null objects to overt objects. Children consistently accepted sentences with overt anaphoric objects and rejected sentences with negative polarity objects, and treated sentences with null objects as fully ambiguous.