Uncivil for Civil Rights: A machine learning and qualitative analysis of incivility in the X-based conversation about Black Lives Matter

Abstract

How does the online disinhibition effect impact communication about the Movement for Black Lives? To answer this question we first use a machine learning algorithm to analyze uncivil language in 1,945,494 tweets from the Black Lives Matter conversation. Mobile users and users in the #BlackLivesMatter hashtag were more likely to use uncivil language than non-mobile users. We also examine a random sample of 993 uncivil tweets in the Movement for Black Lives conversation, and find that many of the users in our qualitative data set employed uncivil language in order to critique racism, and discuss the implications for research using automatic processing to detect uncivil language or hate speech.