【作者简介】张博伦，浙江大学社会学系；周浥莽，中国人民大学社会学理论与方法研究中心

【文章来源】《社会学评论》2026年第2期

【内容提要】计算社会科学在形成过程中存在四条相对独立的知识演化路径，即公理传统、描述传统、预测传统和实验传统。本文从理论与现实关系、进入社会学的路径和方法论困境三个层面梳理计算社会科学的四种传统。在此过程中，本文一方面强调计算社会科学内部的异质性，另一方面论述并试图重构计算社会科学与经典社会学研究方法之间的关系。本文呼吁学界对计算社会科学方法的关注从方法推广转向更严谨审慎的方法整合。

【关键词】计算社会科学 / 公理传统 / 描述传统 / 预测传统 / 实验传统

【全文链接】http://src.ruc.edu.cn/CN/Y2026/V14/I2/5

The Polysemy of“Computation”in Computational Social Science: An Analysis of Four Traditions

Abstract: Computational social science has developed along four relatively independent trajectories of knowledge evolution: the axiomatic tradition, the descriptive tradition, the predictive tradition, and the experimental tradition. We analyze these four traditions across three dimensions: the relationship between theory and reality, pathways of entry into sociology, and methodological dilemmas. In doing so, we highlight the internal heterogeneity of computational social science while also examining and seeking to reconstruct its relationship with classical sociological research methods. This article calls for a shift in scholarly focus from the mere promotion of computational methods toward a more rigorous and cautious integration of three approaches.