社会杂志 ›› 2009, Vol. 29 ›› Issue (4): 148-166.

• 文学研究 • 上一篇    下一篇

再论证伪主义的问题:对尹海洁、庞文商榷的答复

张杨,  博士生   

  1. 芝加哥大学社会学系
  • 出版日期:2009-07-20 发布日期:2009-07-20

On the Problems of Falsificationism Again: A Rejoinder to Yin and Pang’s Critique

  • Online:2009-07-20 Published:2009-07-20

摘要:

本文回应了尹海洁等两位作者对笔者的批评,并再次质疑了证伪主义在社会科学中的适用性。本文指出,由于“迪昂奎因”命题的存在以及社会科学命题的或然性,简单证伪并不可能;而波普尔和尹文所谓的复杂证伪则不可取,因为这种证伪本质上基于归纳法,相对于证实并无优越性和简便性,因而也就失去了波普尔立论的根基。同时,社会科学的统计检验自有其意义,并不需要冠以“证伪”的名号;而“可证伪性”即使存在,但在定量研究中既缺乏操作性,又完全被涵盖于 “解释力”这个概念之中。基于以上理由,“可证伪性”不具备成为社会科学和非科学的划界标准的意义,证伪主义不是社会科学赖以立基的逻辑基础。本文还延伸讨论了证伪、限界条件和发展社会科学经验研究的关系。

关键词:

, 证伪, 可证伪性, 证伪主义, 限界条件

Abstract:

In response to the critique of Yin Haijie and Pang Wen, this article re-accounts for the inapplicability of falsificationism in social sciences. Due to the Duhem-Quine thesis and the probability of the propositions in social sciences, we cannot falsify a theory/proposition by a single exception as Popper’s naive falsificationism puts. Meanwhile, the “complex” falsification proposed by Popper and Yin et al. is in essence based on induction, so it is not superior to or more parsimonious than confirmation, and thus, goes against Popper’s logical starting point for his propositions. Moreover, the value of statistical testing in social sciences does not need to be justified by falsification as falsifiability could hardly be manipulated in practice and it is already totally within the conception of “explanatory power”. Therefore, falsifiability is not a criterion of demarcation for scientific theories from unscientific ones; neither is falsificationism the logical foundation of social sciences. The article also extends the discussion to the relationships between falsification, scope conditions, and developing empirical research in social sciences.

Key words:

, falsifiability, falsification, falsificationism, scope condition