Questions, Relatives, and Minimal Projection

Abstract

This article examines the acquisition of wh-questions and relative clauses in Sesotho, a language with no. wh-movement in either questions or relatives, and in which wh-questioned subjects must be clefted. It shows that, even though children use and understand relative and cleft constructions between the ages of 2 and 3, relative complementizers are frequently missing or they surface in a form that is ambiguously either a subject agreement marker or a head-noun modifier. This raises the possibility that children are treating relative clauses as IP rather than CP structures until sufficient learning of lexical features has taken place. The article concludes with a discussion of Grimshaw's notion of extended projection and minimal projection, showing how it might be adapted to account for the Sesotho findings and extended to acquisition theory more generally.