计算社会科学的生成解释问题  

On Generative Explanation in Computational Social Science

作　　者：袁继红[1] YUAN Ji-hong(Research Center of Intelligent Society and Human Development,Guangdong University of Finance&Economics,Guangzhou 510275)

机构地区：[1]广东财经大学智能社会与人的发展研究中心,广州510275

出　　处：《自然辩证法研究》2020年第4期103-107,共5页Studies in Dialectics of Nature

基　　金：国家社科基金一般项目“计算社会科学的解释研究”(19BZX033)。

摘　　要：计算社会科学的基于能动者建模方法提供了一种特殊的解释路径——生成解释,在方法上,它为自下而上地解释异质能动者的分布式局部互动如何生成宏观规则提供了可能路径;在方法论上,生成解释不是演绎的,它的独特之处在于把生成充分性作为解释的必要条件,是机制解释观的有益补充,为社会科学提供了一种“实验+推演”的实践解释观。The Agent-Based Modeling method of computational social science advances a distinctive explanation approach:generative explanation.In terms of method,it is suited to the bottom-up study of the given macro regularity through the decentralized local interactions of heterogeneous agents.In terms of methodology,generative explanation is not deductive,but it is special in that generative adequacy is taken as the necessary condition of explanation,and it is a beneficial supplement to mechanism explanation,which provides a“experimental+logic”practical explanation approach for social science.

关 键 词：计算社会科学 基于能动者建模 生成解释 演绎 机制解释 

分 类 号：N031[自然科学总论—科学技术哲学]