The Category Of Question And The Generative Mechanisms Of Chinese Interrogatives

The Category of Question is an important grammatical category, thus Chinese interrogatives have always been in the center of the syntactic studies of Chinese. Great achievements have been made in studies of Chinese interrogatives over the past decade, yet still there are disputes on some significant issues. This dissertation reflects upon the past studies, and starts with discussing the Category of Question so as to make it clear on issues as the definition and the grammatical typing of the Category of Question as well as the scope and classification of Chinese interrogatives. Based upon this, it focuses its studies on the generative mechanisms of Chinese interrogatives and the relevant issues within the framework of the Minimalist Program.The main ideas of the dissertation are as the following:(1) The Category of Question has not been clearly defined within linguistic domain and has brought about many misunderstandings epistemically. Traditionally, the Category of Question is regarded as a criterion for classifying interrogatives, and also the pragmatic effect expressed by interrogatives. We take the Category of Question as a purely grammatical concept. It is a non-linear, suprastructural sentential category. The linear realizations and instantiations of the category of question are interrogatives of different types. The relation between the Category of Question and interrogatives is that as abstractness and specifications, category and instantiations. Typologically, the Category of Question can be typed in three layers in grammar, namely, the syntactic layer, the phonetic layer and the lexical layer. The syntactic typing is realized through the interaction of two syntactic positions and three grammatical means. This typing can be summarized as Addition to C, Addition to I, Reduplication in I and Movement to C. The syntactic typing can be typed only but once, which can be summarized as "The Principle of Incompatibility of Syntactic Typing". The lexical typing is preliminary. In some languages the syntactic computation is sensitive to lexical typing, while in some languages it is not. The phonetic typing is complementary to the lexical typing and syntactic typing, and can also be used simultaneously with those two.(2) Interrogatives have the question forms and perform the interrogative function. The criterion for defining the scope of interrogatives should be based on their grammatical forms. Those with question forms and performing the function of asking questions are typical interrogatives, those who have question forms but do not perform the function of asking questions are atypical interrogatives, while those without question forms and functions are not interrogatives at all. Chinese interrogatives can be classified firstly into phonetic interrogatives, syntactic interrogatives and lexical interrogatives based on the way how they type the Category of Question. Phonetic interrogatives are those who type the Category of Question through intonation. Traditionally they are included in Chinese yes-no question. The syntactic interrogatives can be subdivided into 1-typing interrogatives and C-typing interrogatives according to their different syntactic typing positions. I-typing interrogatives are X-not-X question in Mandarin and KVP question in Chinese dialects, and C-typing interrogatives refer to the particle question and "VP-neg" question. The lexical interrogatives are wh-question and disjunctive question. They both mark their Category of Question through the question features of certain words, and their similar syntactic behaviors also provide evidence for grouping them together.(3) Based on Rizzi's (1997) Split CP Hypothesis, we put forward that "ma" actually projects into the head F1 within the ForceP domain. The functional head F1 position holds the grammatical complementizers, i.e the complementizers marking the grammatical categories like the Category of Declaration and the Category of Question. The functional head F2 is occupied by pragmatic complementizers. Superficially, "ma" occurs in the final position of the sentence, or to the right of its complement IP, it evidentially generates to the left of its complement IP. Firstly, Chinese is a head-initial language, and most complementizers occur before their complements. Besides, many head-initial complementizers have gradually grammaticalized into head-final complementizers. Most importantly, the particle "ma" is grammaticalized from the interrogative morpheme before its complement VP. The reason why "ma" occurs to the left of its IP complement is because it moves rightward to its IP due to the expression of focus and the syntactic rhythm.(4) The "X-neg-X" pattern in X-neg-X question is formed through a linear combination of the base form X, an appropriate negative form for X and the reduplication of X in lexicon. If X is a monosyllabic word, an "A-neg-A" form is derived. If X is a disyllabic word, an "AB-neg-AB" form, an "A-neg-AB" form and an "AB-neg-A" form can be derived. The "X-not-X" form has the [+Q] feature. The "A-neg-AB" form has the [+Q] feature in the lexicon and the "A-neg-A" form and the "AB-neg-AB" form has the [+Q] feature in the syntactic computation as their reduplication is realized through the syntactic process. The question morpheme Q is generated within I scope in Chinese X-not-X question, and it has an uninterpretable [+Q] feature. Therefore it demands that the "X-not-X" with the same [+Q] feature move to its position for feature checking. The question morpheme in X-not-X question does not have the need to move forward to C position in LF to acquire its interrogative scope.(5) Chinese disjunctive question and wh-question both do not display overt movement syntactically. No research has been conducted on the generative mechanism of Chinese disjunctive question, the reason being that there is no clear-cut definition for Chinese disjunctive question. We take that the disjunctive question should be defined syntactically. It refers to those in which the disjunctions are joined by the connective with the [+Q] feature so as to provide choices for the hearers. As for the theories about the generative mechanisms of Chinese wh-questions, the covert movement and the movement of interrogative operators, they are both far from satisfactory. We hold that, the connective "haishi" in modern Chinese disjunction question and the wh-words in Chinese wh-questions have the [+Q] feature in the lexicon. In the syntactic computation process, "haishi" and wh-phrases in these two questions percolate their [+Q] features to the head Q, therefore, the uninterpretable [+Q] feature of Q can be deleted and these two questions are thereby generated. In certain syntactic environment, this feature-percolating process can be hindered or restricted.