Hate Speech on Social Media: Uncovering the Language Crevices Where Hate Mongers Hide

In Kenya elections have a history of being divisive. This has become very pronounced in the 2017 disputed presidential poll in which campaigns have become synonymous with hate speech. The hate speech is not limited to campaign trails or to politicians: it has spewed into social media platforms and the perpetrators are supporters of the politicians. Kenya has a law against hate speech, but the said law has led to no convictions and has therefore done little to tame hate mongers on social media platforms. Indeed, the National Integration and Cohesion Commission has considered shutting down social media platforms in the run up to the repeat presidential election (October, 2017) as the only way to ensure that these sites are not used to lead the country to chaos. For some, this would be a violation of fundamental freedom of expression guaranteed in the constitution but the Commission argues, it would be a necessary move to ensure that hate speech on social media platforms does not lead to physical violence. Many have lamented the rise in hate speech and the apparent inability by concerned authorities to reign in the perpetrators. This paper poses the general question: is it totally impossible to deal with hate speech on social media platforms? Drawing data from the legal statutes on hate speech, selected Facebook and Twitter posts as well as online blogs, and interviews with legal experts the paper sought to answer this question. Preliminary finds show there are language crevices in the National Cohesion and Integration Act (2008) make it difficult to prosecute hate speech. In addition, there are linguistic features evident in hate posts that make it difficult to prosecute them as such. Give these findings, the paper suggests a language based criteria can be used to determine hate speech