In search for a common context : opinion polarization in Chinese microblog interactions

Abstract

The polarization of online opinions has been observed in many scenarios. It is deemed as a result of social comparison and selective exposure within users’ homogenous social networks. As different groups are gathered together on social media, how opinions dynamically shift during online debates still remains unclear. Through the lens of social interactionism, this thesis draws upon the analysis of online interactions to explore the dynamic changes of online opinions on Weibo—China’s Twitter equivalent. To examine how Weibo’s online environment impacts on the users’ interactions, the research reconstructs an opinion pool consisting 75,248 comments and 6834 conversations on a national debate on a China-Hong Kong dispute. The study finds Weibo’s information structure and technical configuration has a strong influence on how users debate with each other: The platform’s thread-based message structure segregates the cyber space into multiple publics, whereas the technical design allows users to simultaneously interact with multiple audience group at once. In addition, the 140-character word limit significantly reduces the chance of establishing a unified communication context among Weibo users. Due to these new interactive prerequisites, Weibo’s cyber space becomes highly fragmented. Facing the constantly shifting contexts, users developed three strategies to recontexualize their debates: sharing personal experience, placing environmental elaborations, and seeking group solidarity. These efforts are observed to further widen opinion gaps due to Weibo’s virtual setting, thereby enhancing the polarization of opinions. By focusing on the interplays between the contexts and the debate contents, the research contributes to the existing literature in four main aspects. Firstly, it provides a dynamic perspective to examine the polarization of online opinions. The findings highlight the subtle influence the platform places on the way Weibo users debate with each other, suggesting that the polarization of online opinions is not merely caused by in-group communication. Secondly, it calls for a revisit of the relationship between the context and actors in online settings. By doing so, it offers a more intricate framework of context entrainment comparing with the existing concept of context collapse. It suggests that online users are located in a flux of fragmented contexts instead of merely collapsing into a convergence of social networks. Thirdly, the research provides a detailed approach to look at the shift of the opinion landscape of Chinese Internet. Lastly, this research also presents an original data-mining method that could help researchers to efficiently gather data on Weibo and other static webpages.