This seminar examines ongoing research on a subset of hate speech, “dangerous speech,” that has a special capacity to catalyse violence. It describes an innovative framework to identify such speech, which was adapted for use in the yearlong Umati project – in which Kenyan monitors scrubbed their country’s online spaces, for hate speech and dangerous speech. They assembled a remarkable dataset, which informed attempts to diminish the force of dangerous speech. The seminar will also explore how a similar approach can be adapted to other normative climates.

For more information see: