Variables and Values in Children’s Early Word-combinations

A model of syntactic development proposes that children’s very first word-combinations are already generated via productive rules that express in syntactic form the relation between a predicate word and its semantic argument. An alternative hypothesis is that they learn frozen chunks. In Study 1 we analyzed a large sample of young children’s early two-word sentences comprising of verbs with direct objects. A majority of objects were generated by pronouns but a third of children’s sentences used bare common nouns as objects. We checked parents’ twoword long sentences of verbs with objects and found almost no bare common nouns. Children cannot have copied sentences with bare noun objects from parents’ two-word long sentences as frozen chunks. In Study 2 we raised the possibility that children’s early sentences with bare nouns are rote-learned ‘telegraphic speech’, acquired as unanalyzed frozen chunks from longer input sentences due to perceptual problem to hear the unstressed determiners. To test this explanation, we tested the children’s speech corpus for evidence that they avoid determiners in their word-combinations. The results showed that they do not; in fact they generate very many determiner-common noun combinations as two-word utterances. The findings suggest that children produce their early word-combinations of the core-grammar type by a productive rule that maps the predicate-argument relations of verbs and their semantic arguments to headdependent syntax, and not as frozen word-combinations. Children mostly learn to use indexical expressions such as pronouns to express the variable semantic arguments of verbs as context dependent; they also employ bare common nouns to express specific values of the arguments. The earliest word-combinations demonstrate that children understand that syntax is built on the predicate-argument relations of words and use this insight to produce their early sentences.