证伪：社会科学研究的可能与必然 兼与张杨商榷

本文针对张杨《证伪在社会科学中可能吗？》一文展开争鸣与论述，指出证伪是理论检验的必然结果之一，相对于证实而言，证伪更有效、更可靠。本文还就张杨文中对波普尔证伪理论的误解做了澄清，指出波普尔的证伪并非“单个反例证伪”，证伪、可证伪性、证伪主义三者之间也存在着明显的区别，不能加以混淆。文章就张文提出的证伪在社会科学中面临的四个挑战做出一一回应：（1）“迪昂奎因命题”不能成为证伪的诘问；（2）穆勒命题不能挑战证伪主义；（3）统计陈述不是概率陈述；（4）限界条件与证伪的结合使社会科学的理论更加科学。

Falsification: Likelihood and Necessity in Social Science Research-A Discussion with Zhang Yang

Challenging Zhang Yang’s arguments in his article Is Falsification Possible in Social Science?， this paper contends that falsification is one of the inevitable results in theory testing and that it is more effective and reliable than confirmation. This paper tries to clarify the misunderstandings of Popper’s falsificationism in Zhang’s paper. Popper’s falsification is not to “falsify by a single counter example，” and there exist obvious differences among falsification， falsifiability， and falsificationism， which should not be obscured. The current paper also responds to the four challenges that falsification has to face in social sciences as pointed out by Zhang： the Duhem Quine thesis can’t hamper falsification； Mill’s proposition can’t challenge falsificationism； statistical statements are not probability statements； and the theory of scope conditions combined with falsification will make the theories in social sciences more scientific.