Researchers have long noted that a given child will variably produce a given morpheme. This has been typically attributed to impoverished syntactic or semantic representations. This paper reviews a variety of evidence from several languages showing that much of this early variability is due to constraints on children's early prosodic representations. In accord with Gerken's (1996) experimental findings, the authors show that 2-year-olds, in spontaneous speech, are more likely to produce certain grammatical morphemes in prosodically licensed (unmarked) contexts, where determiners can be prosodified as part of a disyllabic (metrical) foot, and 3rd person singular morphemes can be prosodified as a simple coda consonant. These findings suggest that children may have earlier grammatical competence than often assumed. If so, this raises concerns about theories of syntactic acquisition, as well as for the nature of the experimental design.
Proceedings of the 2nd Conference on Generative Approaches to Language Acquisition North America (GALANA)
edited by Alyona Belikova, Luisa Meroni, and Mari Umeda
Table of contents