以国内外学者对副词“都”的语义功能的探索为例,倡导一种“由问题推动”的语法研究路线:以理论的眼光观察语言现象、提出明确的问题,然后通过分析相关的语言事实来论证和解决所提出的问题,并发展出新的理论原则和分析方法。首先通过提出关于“都”的语义的各种问题,引出关于“都”的语义功能的“总括”学说、全称量词理论、分配算子理论、加合算子理论,指出它们各自在理论和事实方面的困难;然后,提出一种兼擅分配学说和加合学说的综合性理论——由加合导出分配;最后,通过研究案例说明:“问题驱动”的研究策略可以兼具“数据驱动”和“理论驱动”两种研究策略的优点。
Taking the exploration on the semantic function of the Mandarin adverb dou as an example, this paper advocates a question-driven approach to grammatical researches, which means investigating linguistic phenomena in perspective of certain theory, raising some explicit questions, and further answering the questions by analyzing the related linguistic facts with logic proving, and at last, developing a new theory or methodology. By means of raising various interesting questions about the meaning of dou, the paper firstly reviews some main theories and proposals on the semantic function of dou, i. e., totalization theory, universal quantifier, distribution operator, sum operator, and points out their inadequacy with respect to logic theory and linguistic fact. Then we propose an integration theory of sum operator and distribution operator, which means dou operates on a set of minimal events, it causes the semantic effect of distributivity. Finally, through the case study of the meaning ofdou, we demonstrate that the question-driven approach has the advantage of both data-driven approach and theory-driven approach.