Many theorists have dismissed a priori the idea that distributional information could play a significant role in syntactic category acquisition. We demonstrate empirically that such information provides a powerful cue to syntactic category membership, which can be exploited by a variety of simple, psychologically plausible mechanisms. We present a range of results using a large corpus of child‐directed speech and explore their psychological implications. While our results show that a considerable amount of information concerning the syntactic categories can be obtained (...) from distributional information alone, we stress that many other sources of information may also be potential contributors to the identification of syntactic classes. (shrink)

Chomsky proposes a reformulation of the theory of transformational generative grammar that takes recent developments in the descriptive analysis of particular ...

We present a series of three analyses of young children's linguistic input to determine the distributional information it could plausibly offer to the process of grammatical category learning. Each analysis was conducted on four separate corpora from the CHILDES database (MacWhinney, 2000) of speech directed to children under 2;5. We showthat, in accord with other findings, a distributional analysis which categorizeswords based on their co‐occurrence patterns with surroundingwords successfully categorizes the majority of nouns and verbs. In Analyses 2 and 3, (...) we attempt to make our analyses more closely relevant to natural language acquisition by adopting more realistic assumptions about howyoung children represent their input. In Analysis 2, we limit the distributional context by imposing phrase structure boundaries, and find that categorization improves even beyond that obtained from less limited contexts. In Analysis 3, we reduce the representation of input elements which young children might not fully process and we find that categorization is not adversely affected: Although noun categorization is worse than in Analyses 1 and 2, it is still good; and verb categorization actually improves. Overall, successful categorization of nouns and verbs is maintained across all analyses. These results provide promising support for theories of grammatical category formation involving distributional analysis, as long as these analyses are combined with appropriate assumptions about the child learner's computational biases and capabilities. (shrink)

This work offers a comprehensive picture of the dynamic natures of language and consciousness that will interest linguists, psychologists, literary scholars,...

The research reported here is a model of language acquisition that addresses how the child maps the input data into output. The project is presented as a response to certain learnability constraints on theories in developmental psycholinguistics. One of these constraints is the continuity of categories across development. Where other learnability models assume innate perception of syntactic categories , this simulation assumes that they are learned. ;The subjects were 11 mother-child dyads, who were videotaped in a laboratory playroom once a (...) month from the time the children were 8 months old, until they were about 28 months old. The input data were obtained from mother's speech during three observations which were milestones in the children's language acquisition: the first word, vocabulary spurt , and the point at which the mean length of utterance surpassed 1.5 words. The output data were obtained from the child's speech during two observations: one was the third milestone and the second was the month immediately following. ;The child's primary data were coded and then analyzed by a computer program. The input consisted of language paired with situational representations. The categories for the situation representation assume a nominalist ontology and are, therefore, "concrete" categories. The categories derive from world, rather than linguistic, knowledge and are a subset of Jackendoff's conceptual semantic categories. The subsetting was motivated by concerns within the casual theory of meaning. ;The output of the simulation consisted of verb general and verb specific sentence frames. Both types of frames in the children's speech were subsets of the mothers' input, meaning that what the child learned was available in the mother's speech. These frames fulfill the function of canonical sentences, which, in several learnability-theoretic accounts, are presumed innate. However, they do not, themselves, account for the presence of abstract objects in the child's later syntactic projections. The verb specific argument structures are shown to be adequate to this task. Harrisian transformations on sets of sentence frames , were invoked to account for the abstract objects. (shrink)