This thesis seeks to specify empirical differences between two types of rhetoric thought to contribute to the onset of genocide: dehumanization and toxification. It utilizes radio transcripts from the Rwandan Genocide to test two propositions: that toxification and dehumanization are empirically distinguishable, and that toxification contributes to the onset and/or intensification of killings in a genocidal context. Results indicate that there are empirically demonstrable differences between dehumanization and toxification, but toxification does not contribute to the onset or intensification of genocide. Instead, the Rwandan case indicates toxification may be utilized as an attempt to motivate latent perpetrators to participate and justify the actions of those already participating in the genocide, as well as to attempt to maintain power in the face of perceived loss. This thesis contributes to the literature on dehumanization and the uses of language in genocide.