Using grammar as a shield in threatening, hateful online communications
Tanya Karoli Christensen
 
To most people, threats and hate speech are considered harmful and disruptive forms of communication that should be discouraged and, if possible, prosecuted as illegal. This can be seen in the often immediate calling out of hateful and threatening messages online, e.g., through hashtags labeling a previous post or news headline as #hatespeech or urging other users to report a perceived threat to the police.
In the many-to-many social media interactions on, e.g., Facebook and Twitter, where a broader audience is both omnipresent and heterogeneous, we rarely see direct, unambiguous threats where the sender takes responsibility for a future harm directed at a victim (such as the archetypical “I’m gonna kill you”). Instead, what we find is a multitude of other ways of expressing ill wishes for others. Interestingly, it is often not the type of harm that is veiled but rather the responsibility for carrying it out.
Based on a unique data set of reports of hate speech in Denmark, provided by the Center for Prevention of Exclusion, I examine the grammatical devices used in racially discriminatory Facebook comments to distance senders from responsibility for the harm they envisage for their targets. Frequent shifts away from a 1st person agentive subject include using the passive voice, modals, verbs of hope or desire, and imperative clauses. I hope to engage you in a discussion of how these shifts affect the linguistic analysis of these comments as threats.
I must warn that the data material contains highly disturbing content. While I seek to protect viewers from the worst parts, the analysis does require a look at language that expresses harm to others. If this works as a trigger for you, please prioritize your mental wellbeing above attending this talk.
Needless to say, I do not subscribe to the views represented in the data.