The Extended Projection Principle And EPP Effects In Mandarin Chinese

This dissertation aims for an explanatory study of the Extended Projection Principle (EPP) and EPP effects in Mandarin Chinese in the framework of the Minimalist Program (MP). So-called EPP effects refer to obligatory raising of VP-internal arguments to the Specifier position of a functional category. It is generally believed that the existence of EPP effects in natural languages provides indirect evidence for the establishment of the EPP as a UG principle.Ever since it was formulated in early1980s, the EPP has been a significant subject in generative grammar. It is only with a coherent theory of the EPP can we gain explanatory adequacy over empirical data and achieve a high degree of internal consistency among modules of the UG. Therefore, EPP studies are of great theoretical significance as well as empirical value.In the GB-model, the EPP was formulated as a requirement that sentences must have a subject. Lacking of explanatory adequacy, the EPP at this stage was descriptive in nature, due to which its status was challenged. Since early1990s, the EPP has been re-formulated in the light of the Minimalist Program. EPP effects have been attributed to an uninterpretable D-feature on T, which must be checked by a raising argument or an expletive element. Compared with its GB precedent, this checking theory of the EPP has made significant improvement as EPP effects have been unified to a feature-based system of syntactic operation. In Chomsky (2000) and afterforth works, it has been proposed that the EPP as a requirement for a filled-specificer may apply to all core functional categories (CFCs)-Light Verb Categories (vP), Tense Categories (TP) and Complementiser Categories (CP). In the latest Phase-theory, EPP-feature as a term is replaced by edge-feature, yet related studies still center on obligatory argument-raising, expletive-insertion and other EPP-effects. The discussions concentrate on whether the EPP is an independent UG principle and what in nature the EPP-feature is.However, the nature of the EPP feature has never been unanimously agreed on. Proposals for the nature of the EPP (-feature) vary from one to another, ranging from syntactic, phonological, morphological features to interface requirement. There are also proposals that the EPP be eliminated from UG due to its redundancy with other principles, for example, Case and theta-theory. Although the proposals so far are all empirically supported, no one covers all the facts. That is why the EPP looks much like the story of Six Blind Men and the Elephant.Meanwhile, there is an empirical issue concerning the EPP-universality. The issue can be understood from two aspects:one is whether it applies to all languages; the other is whether it applies to all functional categories. Answers to these two questions calls for cross-linguistic studies. In the last decade, investigations of EPP property in European, Asian and African languages have collected a large amount of data, some supporting the EPP whereas others not. Recently, some researchers adopting the common practice advocated by Chomsky that "assume languages to be uniform in the absence of compelling evidence to the contrary" start from a null hypothesis that EPP applies to all languages, testing the hypothesis against particular languages one by one. That is also the paradigm adopted in the present dissertation.Although a number of studies have been carried out in languages other than English, few studies have been conducted of Mandarin Chinese (MC). Linguists hold different opinions on whether MC is an EPP language. Advocates argue that Chinese sentences all have a subject, overt or covert; whereas opponents cite non-subject sentences as a challenge. It must be pointed out that it is problematic to reduce the EPP-property to the presence of subjects since a GB-based phrase structure rules proved to be descriptive. In the framework of MP, answers to that question should be based on EPP-effects observed in MC. That is, focus should be on whether MC exhibits EPP effects and what triggers the effects. On the assumption of Last Resort Principle, the EPP could be established as a piece of UG principle only if other motivations are dismissed from triggering the EPP. Considering that MC is poor in inflectional morphology, EPP effects reported in MC will enrich EPP-theories which originally evolve from Indo-European languages, particularly help exclude redundancies between the EPP and inflectional morphologies long-suspected in English. Two issues will be discussed in the present dissertation. One is the nature of the EPP; the other is EPP effects in MC.Working within the Minimalist Program (MP) in its latest versions, the author puts forward an original hypothesis about the nature of the EPP-the Fun-Hypothesis. Key points of the Fun-Hypothesis are as follow:a. The EPP is a selectional feature [Fun] on a functional head which projects an independent projection.b. As a privative feature,[Fun] is universal to functional heads cross-linguistically.c. The checking of [Fun] needs to be accomplished via merging with a constituent bearing [Sub].Compared with standard theories of the EPP which rely on Spec-head relation to account for EPP effects, the Fun-hypothesis adopts a head-complement relation which is believed to be more fundamental than Spec-head. In the case of CHL, this helps reduce the burden of computation and achieve a higher degree of theory-internal consistency, conforming to Chomsky’s idea of language being optimal.In addition, EPP effects have been shown to widely exist in MC with respect to vP-domain, TP-domian and CP-domain via a close study on the distribution of arguments in certain structures. Particularly, the redundancy of the EPP and Case is dismissed in the case of MC which is an inflectional poverty language. This has provided indirect evidence for the cross-linguistic universality of the EPP.It has been shown that analysis of the EPP effects from the perspective of the Fun-Hypothesis has some advantages with regard to empirical coverage. A new perspective has been provided via the Fun-Hypothesis.Contributions of the present dissertation can be seen from two aspects. First, the Fun-Hypothesis proposed in the dissertation provides a new perspective upon the EPP. If the hypothesis is on the right track, it helps reduce the burden of computation and achieve a higher degree of theory-internal consistency. Second, analyses of EPP effects in MC carried out in the present dissertation present empirical data supporting the null hypothesis that the EPP applies to MC. Limitations of the study is that there is a lack of ample cross-linguistic evidence for the hypothesis. The format of the dissertation is as follows:Chapter One provides a brief introduction of research background and major proposals; Chapter Two reviews major approaches to and key studies of the EPP, along with cross-linguistic data which is convincingly cited as empirical evidence. Chapter Three illustrates the major proposal and theoretical framework under which the syntactic derivation in this work is to be computed; Chapter Four concentrates on EPP effects exhibited by functional categories in MC; Chapter Five is the conclusion.