Search

The acquisition of complement clause constructions in English, like in other languages, has received relatively little attention from researchers, even though the development of these complex sentence structures can shed light on the language acquisition process, and can hence inform us about the accurateness of suggestions put forward by different theoretical stances. The previous corpus research that has been conducted in this field has been based on relatively sparsely collected datasets, while a large majority of experimental studies have focused on sentences with to-infinitive clauses as control structures often ignoring other aspects of development and other construction types. Due to the need for a detailed investigation, the present thesis first investigated the development of all finite and non-finite complement clause constructions in one English-speaking child's densely collected data (5h/week). This analysis showed that many complement clause constructions appear within a relatively short period of time during the latter half of the third year of life. We suggest that input, cognitive/functional salience, and semantic and pragmatic factors are likely to play an important role in the development and appearance of these constructions. The effect of input was then investigated further by looking at (1) complementizer omission errors produced with certain non-finite complement clauses, and (2) ACC-for- NOM lpsg sentence subject errors in 14-17 children's speech. These errors were found to be tightly linked to the competition between different forms in the input that the children were exposed to. The complementizer omissions were shown to be related to children learning one- (want-X, going-X) and two-word chunks (want-to-X, going-to-X), which were competing for output. The ACC-for-NOM errors were linked to the input and competition of I-verb sequences in simple sentences and me-verb sequences in non-finite complement clause constructions (e. g. Let me do that). Finally, the effect of input was investigated through a narrower window by analysing how input in the immediate discourse context affects one child's production of that- and to-infinitive complement constructions in naturalistic conversation between 2;9 - 3;11. The two constructions were found to differ in the amounts of lexical and structural overlap with adult utterances in prior discourse, suggesting that (a) they developed at a different pace despite emerging very closely together in the child's speech and (b) close-proximity-input may aid the production of difficult sentence structures early in development. The present research largely supports the Constructivist view of language acquisition. In particular, it adds to the previous research suggesting that input plays a major role in children's adult and non-adult-like language knowledge.