The emergence of syntactic categories: Evidence from the acquisition of Japanese.

摘要

This dissertation specifically addressed three issues pertaining to the emergence of syntactic categories and examined some proposed hypotheses in the light of data from the acquisition of Japanese.;The first issue was how children identify words in terms of grammatical categories and functions. Pinker's bootstrapping hypothesis, O'Grady's general nativist theory, and Braine's assimilation-accommodation theory were directly examined. The data indicates that verbal nouns--not regular verbs--were initially used to encode actions/changes of state by children. Thus, the early emergence of verbal nouns found in the data created a problem for all the theories in one way or another.;The second issue was whether nouns are acquired before verbs. The data points toward the early primacy of nouns. However, why nouns were acquired before verbs was not clear because of the early emergence of verbal nouns. Meaning-based hypotheses per se or the explanation based on the nature of input alone cannot account for the early emergence of regular nouns and verbal nouns and the late emergence of verbs: verbal nouns emerged much earlier than verbs despite the fact that both have the same type of meanings, and verbs were more frequently used in the input. Thus, the explanation based on morphological complexity needs to be reconsidered, since like regular nouns but unlike verbs, verbal nouns are associated with a lesser degree of morphological complexity.;Finally, for the issue of the acquisition of lexical and functional categories, the maturational and the semantic hypotheses were examined. Since there is no distinction between the two categories in Japanese, these two theories make different predictions for the timing of the emergence of Japanese determiner-like elements, which roughly correspond to English Ds. The data indicates that these elements were acquired late. The late acquisition of these elements can only be attributed to the relatively abstract semantics of these elements (as they are not functional categories), which makes it unnecessary to attribute the late acquisition of English determiners to their categorial status as functional categories. Thus, the data suggests that the semantic theory seems to be preferred.