PIC And The Acquisition Of Unaccusatives, Ba-Constructions And Passives In Mandarin-Speaking Children

The present research consists of two large sample experiments on the acquisition of unaccusatives, object-raising Ba-constructions and passives of Mandarin-speaking children. It is found that there is an apparent delay in the acquisition of passives of Mandarin Chinese children as is observed in many other languages. As is shown in our data, a very high percentage of the subject children in 3 year olds in our experiments do not have difficulty in the comprehension of Ba-constructions as compared with their low percentage of them encounter difficulty in the comprehension of passives, and the percentage of those who are able to produce passives is much lower than that of those who are able to produce Ba-constructions in all age groups. Even up to the age of 5 and up, production of passives is observed to be problematic at a higher percentage than that of Ba-constructions. There seems to be no problems at all for all the age groups in both comprehension and production of unaccusatives, even for the younger age group of 3 year olds. Our experiments lend a firm empirical support to the alleged universality of the delay of passives.On the basis of Chomsky's (2001) Phase Impenetrability Condition (PIC), Wexler (2004,2007) proposed his Universal Phase Requirement (UPR) to account for the universal delay of passives. The results in our research suggest a new definition of phase for PIC and UPR, which can be stated as follows: a vP with an externally merged Spec-agent is a strong phase but a vP without externally merged Spec-agent is not. Under this specific definition of phase, with an appropriate syntactic analysis of the constructions involved, PIC and UPR can well account for the acquisition fact of delay in passives. When the movement of the internal argument in passives crosses a strong phase (vP) with the externally merged Spec-agent appearing after the functional item Bei, it violates the PIC. Since the strong vP phase is weakened or defective for adults as assumed in Chomsky's PIC but remains strong for young children as assumed in Wexler's UPR, the reason why young Mandarin speaking children perform poorly with passives but pretty well with unaccusatives and Ba-construction is in order. The derivation of unaccusatives and Ba-construction are PIC and UPR irrelevant.The theoretical contribution of the present project is with the specification of vP phase in terms of the availability of Spec-agent following a stricter implementation of Merge: when no agent is present as in passives, there will be no Spec position since there is no such a position before agent is merged externally. As is implied in the present research, the so-called delay of passives in child language acquisition is not an abnormal process but the natural steps toward maturation. In fact, the delay of passives suggest that there is an acquisition order or implicational relation at least between passives and unaccusatives and Ba-constructions in Mandarin Chinese– the acquisition of Ba-constructions and unaccusatives proceeds that of passives, that is, the acquisition of passives depends on that of unaccusatives and Ba-constructions. The successful acquisition of passives implies the successful acquisition of unaccusatives and Ba-constructions.The other major contributions of the thesis are: (A) as is indicated in our data, the familiar dichotomy between long passives and short passives is not empirically well-grounded. The functional item Bei in the so-called long passives is not identical with the Bei in the so-called short passives since the Bei in long passives can alternate with Gei, Rang, Jiao whereas the Bei in short passives can only alternate with Gei. Thus, it is not reasonable to claim that short-passives are the shorthand form of long passives. Short passives are in fact unaccusatives with the optional Gei/Bei as the detransitivizer. (B) there seems to be some cases in our data showing that in a language without inflections like Mandarin Chinese, the Optional Infinitive Stage (OIS) as proposed in Wexler (1990, 1992, 1994, 1998), Kallestinova (2007) and confirmed in many inflectional languages can also be observed. Some of the young children in the age group of 3 year olds do not use the lexical functional tense marker le for tensed verbs as adults do. (C) The analysis of the data in the experiments favors the syntactic analysis of Bei in Huang (1999) and that of Ba in Tian (2003) and Shi (1997).