From abhorrent to acceptable: norms, standards, and negativity in online communities

In recent years, toxic online communities (characterised by prejudices, negativity and aggression) have attracted increasing concern over their potential to foster and amplify antisocial sentiments, giving rise to expressions and acts of intolerance, hate and violence offline. Across four studies, this thesis explores how forms of negativity can become normalised, acceptable and perpetuated in these online spaces, informed by Phillips and Milner’s (2017) work on ambivalence. Integrating results from a longitudinal interview study (N=44, 12-76yrs) with previous research, Study One presents a taxonomy of processes contributing to the emergence and maintenance of norms and cultures in online communities. This study provides insight into how toxic communities form, propagate and resist change. Study Two is a second interview study (N=16, 18-58yrs) exploring how individuals perceive and reason about diverse forms of online negativity. In examining the factors considered when individuals evaluate online interactions, the study identifies mechanisms by which antisocial behaviours can be reconstrued as normal, benign and acceptable. The final two studies are mixed-methods experiments testing the effects of factors identified in Study Two as influencing appraisals of specific types of online aggression. Focusing on celebrity hate and critique on Twitter, Study Three (N=140, 17-57yrs) indicates that attacks on celebrities are perceived as less serious than those targeting non-celebrities, as they are considered common and therefore expected and minimally harmful. Attacks on targets’ behaviour are similarly considered better than weight- and appearance-based attacks, as victim blaming allows individuals to shift responsibility to targets and justify negativity. Finally, Study Four (N=117, 18-72yrs) examines the influences of humour and supportive community norms in neutralising aggression on Reddit. These factors evoke more positive perceptions and responses from witnesses, who conclude that such behaviours are not malicious, less hurtful, and therefore more acceptable. Together, these studies provide insights about how behaviours typically considered antisocial can become normalised and condoned by individuals, and how these favourable attitudes can potentially spread and perpetuate toxic cultures in online communities.