“事实”如何胜于“雄辩”?——基于语料库量化之“功、过”考察

How to Make‘Facts’Speak Louder than ‘Words’——Based on the‘merits and demerits’of corpus quantification

近年来，语料库作为一种教学和科研的“有效”定量研究方法，受到人们极大关注，因为它或多或少解决了乔式理论中语言“直觉”意识缺陷，满足了“观察充分”和“描写充分”两个标准，减少了材料选择的主观性，提升了客观性和科学性。然而从语料库建构角度看，它并非灵丹妙药，因为它建立在人为选料基础之上，不可避免地存在与生俱来的问题，如代表性、规模性和真实性等。因此，笔者认为“事实胜于雄辩”并非永远不变的真理，关键要看何种“事实”、以何种方式建构或使用“事实”。只有定量与定性相结合，有血有肉的事实（数据）才更具说服力、才更有用。

In the past few years, corpus as an ＇ effective＇ quantitative approach for teaching and research, has at- tracted people＇ s great attention, for it has somehow tackled verbal ＇ intuition＇ default of Chomsky＇ s theory and met two criteria of linguistic theory evaluation： observational adequacy and descriptional adequacy, which decreases subjectivity and increases objectivity and scientificalness of material selection. However, from the perspective of corpus construction, it is not panacea, for it is based on human selection, which cannot avoid some inborn issues, such as representativeness, scale, naturality and the like. Therefore, the authors hereby argue that ＇ facts＇ -speak-louder-than-＇ words＇ is not the case forever, and what is most important is what kind of ＇ facts＇ , how to construct or use them. Only when quantitative and qualitative researches integrated well can the organic ＇facts＇ （quantifying evidences ） be more convincing and useful.