网络左翼的三重面相:基于个案观察和大数据的探索性研究

本文在质性观察和归纳的基础上提出一个由“思想理论资源”与“立场取向”两个维度构成的理论模型，用以建构关于网络左翼思潮的类型学。在此基础上，本文使用个案观察加大数据文本分析的混合方法对三种网络左翼思潮进行研究，从经验上辨识与类型学一致的左翼社会思潮，即“国家主义左翼思潮”“民粹主义左翼思潮”与“自由主义左翼思潮”。此外，我们利用监督性机器学习技术与社群侦测技术，辨识各个网络左翼派别的社群结构，并分析它们彼此之间的网络互动与社群结构的演进。研究发现，随着时间的推移，“自由主义左派”逐渐式微，不再构成相对独立的互动社群；“民粹主义左派”与“国家主义左派”之间存在更密切的关系，但两者间裂痕的扩大也带来网络冲突的增加。

Three Faces of the Online Leftists: An Exploratory Study Based on Case Observation and Big Data Analysis

Leftist social thought is not a unitary and coherent system of thought and left-wingers are made up of divergent groups. In this study, we propose a theoretical typology of two dimensions of theoretical resource and ideological orientation to analyze left-wing social thought in online space. A combination of case observations and big data analyses of Weibo tweets is applied to investigate the three types of leftist social thought online, identified as state-centered leftism, populist leftism, and liberal leftism. State-centered leftism features a strong support of the state and the current regime and a negative attitude towards the West. Populist leftism is characterized by the unequivocal affirmation of revolutionary legacies and support of grassroots movements for disadvantaged population. Liberal leftism maintains a grassroots position and a decided affirmation of individual rights. In addition, supervised machine learning and social network analysis techniques are used to identify online communities of the three types of leftist social thought and analyze the interaction patterns within and across the communities as well as the evolution of community structures. This study found that during the period of 2012-2014,the camp of liberal leftist gradually decreased in numbers and the corresponding communities dissolved while the interaction between populist leftist and state-centered leftist camps intensified and the opinion differences between the two increased online confrontations. This article demonstrates that the mixed approach of combining traditional methods with big data analysis has enormous potential in the sub-discipline of digital sociology.