One of the long-standing issues in the study of language acquisition has centered about what is 'innate' and what must be 'learned'. Much of this debate has focused on structures at the syntax/semantics interface, dealing specifically with how children acquire the argument structure of verbs. Some scholars have proposed that the semantics of verbs is especially salient for children, and propels the acquisition of the syntactic structures with which they are used. Others have proposed that the syntactic frames in which verbs appear provide children with essential evidence for constructing the meanings of verbs. Still other researchers have proposed that verb meanings and their syntactic frames are acquired item by item as unanalyzed 'constructions'. Interestingly, there seems to be some support for each of these positions, possibly because of the relatively conservative word order of English, the language on which the majority of these studies have been based. The purpose of the project is to explore these issues more fully using evidence from the Bantu language Sesotho, where word order is flexible and productive verbal morphology signals an increase or decrease in the number of arguments a verb may take. Bantu applicative constructions (also known as 'benefactive' or 'double object' constructions) present a particularly interesting area for exploring the semantic, syntactic, morphological, and lexical contributions to language learning. The project will use multiple sources of evidence, including data from children's longitudinal spontaneous speech productions and performance on experimental tasks. Preliminary evidence from the spontaneous speech of 2-3-year-olds suggests that children are conservative 'lexical' learners, making few errors of commission. Applicative verbs are, however, used in a number of syntactic frames, mediating against a strict constructionist' approach. Nonetheless, certain syntactic frames, such as 'double object' constructions with both Benefactive and Theme object NPs, never occur in children's spontaneous speech, nor in the adult speech directed toward them. The data from younger children will be complemented with an examination of older children's use of applicatives in spontaneous speech, where errors of commission are expected. These results will be compared with results from cross-sectional experimental studies designed to test children's use of the applicative with novel verbs from different verb classes, and to test their developing knowledge of word order restrictions in 'double object' constructions.