The acquisition of double-gapped relatives in Chinese: Grammar and processing

This dissertation is an acquisition study of double-gap relatives in Chinese, developed under the thesis of subject-object asymmetry as postulated in Huang (1984). Two sets of act-out experiments were conducted. The first experiment tested covert-binder relatives, i.e. double gap relatives with the binder of their non-head EC unmentioned in the sentence. This type of relative was presented in two contexts: pragmatically neutral vs. biased conditions. Assuming that in the absence of pragmatic bias there is a requirement for the head of a double gap relative to be construed as the object of the relative, it was speculated that the provision of a topic NP might produce a swing towards the head of the relative being construed as the relative subject. Under this speculation, it was predicted that there would be a higher proportion of subject interpretation of the relative head in the pragmatically enriched condition than in the neutral condition. The second experiment is a follow-up, testing overt-binder relative, i.e. double gap relatives with either an overt matrix subject or topic as a potential binder of their non-head EC of the relative. The prediction here was that while a subject interpretation of the head for a subject-binder relative should be linguistically blocked, the head of a topic-binder relative was allowed to be interpreted either as the null subject or as the null object of the relative. The main result is that there is a general preference for construing the head as the null object for both covert- and overt-binder relatives. An additional finding is that the younger children, but not the older children, show some sensitivity to context, giving more object construals under the enriched condition. Huang's account is not supported by the data. The general preference for the head as null object is explained in terms of an independently attested principle of sentence processing: empty categories are always identified as soon as possible in on line coreference assignment (e.g. Nicol & Swinney, 1989; Gibson, 1998). Young children's sensitivity to the enriched context calls for future work in the field of acquisition and processing.