UC Berkeley

By studying the social media apex court news in Taiwan and the United States, this dissertation aims at providing new evidence to the institutional legitimacy literature, specifically, whether the existing legitimation such as legal language is still relevant, and to what extent do competing themes of political wrangling challenge the procedural fairness models of judicial legitimacy in the current social media culture.

The dissertation consists of two main parts. First, drawing upon topic modeling techniques, it will review the topics on social media news outlets to see how the traditional legitimacy factors —the powerful symbols of justice and procedural fairness come into play. Both Twitter and Facebook data confirm the procedural fairness models are still relevant. Nevertheless, the justice confirmation procedures in the United States involve considerably more political wrangling themes.

Second, this dissertation studies the judicial legitimacy affect in social media news coverage. It tests whether the judicial language styles in networked news coverage are associated with positive affect more often than with negative affect, and finds quite opposite results in Taiwan and the United States.

The dissertation thus intends to make the following contributions. Methodologically, it ties a variety of tools, including topic modeling and sentiment analysis, to the measurement of legal language, positivity bias and judicial legitimacy affects. In so doing, it debunks the processes of contemporary judicial legitimation, and lays out the vital communication process of judicial information.