Balancing the Abstraction and Empirics of Modeling: A Theoretical Approach to Social Science

理论广泛存在于社会科学研究的各个领域，不同的理论家还时常会就如何发展理论形成争议，其往往涉及理论抽象性与现实性的权衡问题。理论应该有包括输入变量、控制变量、基础性逻辑机制、输出变量以及“环境或外部约束条件”在内的五个基本要素。尽管对于理论构成要素及其现实性的理解有所不同，但弗里德曼所强调的“预测检验法”与科斯所倡导的“要素现实性考察”其实是互相补充而非对立矛盾的关系，它们应该有机地纳入理论建构及发展活动之中。在建模越来越容易的当今时代，基础性逻辑机制展现出现实化的趋势，理论“抽象性与现实性的权衡”问题越发需要在“五个要素”的基础上加以认真对待、系统反思。把握好“五个要素”的理论观，将有助于研究者更好地进行理论的整合与创新，加深对人类话语体系及其社会实践的理解。

How to develop a theory generates much academic debates, which involves the issue of balancing the abstraction and reality of a theoretical model. Five fundamental elements, including input variables, control variables, basic logic mechanism, output variables, and the environment or external constraints should be taken into consideration in a theory. Although holding different opinions on the elements of a theory and the requirements of reality for those elements, M. Friedman’s emphasis on “disprove or fail to disprove predictions” and R. H. Coase’s proposition on “assumptions should be realistic as well as tractable” do not contradict each other, but are in a complementary relation. Both should be integrated into the theory construction and development. As model construction is getting easier than before, basic logic mechanisms tend to correspond better with the real world. Thus, the issue of balancing the abstraction and reality of a model should be taken more seriously based on the “five elements” approach. It can be helpful for researchers to advance the work of theoretical integration and creation.

王水雄、梁鹏飞

本文受到国家社科基金重大项目“网络社会的结构变迁与演化趋势研究”（项目号：15ZDA045）的资助。

欢迎阅读《社会学研究》！您是该文第3530位读者！