Dimensions of Abusive Language on Twitter

In this paper, we use a new categorical form of multidimensional register anal- ysis to identify the main dimensions of functional linguistic variation in a cor- pus of abusive language, consisting of racist and sexist Tweets. By analysing the use of a wide variety of parts-of- speech and grammatical constructions, as well as various features related to Twit- ter and computer-mediated communica- tion, we discover three dimensions of lin- guistic variation in this corpus, which we interpret as being related to the degree of interactive, antagonistic and attitudinal language exhibited by individual Tweets. We then demonstrate that there is a signif- icant functional difference between racist and sexist Tweets, with sexists Tweets tending to be more interactive and attitu- dinal than racist Tweets.