Countering Dangerous Speech Around the World

The Dangerous Speech ideas are meant as tools to help reduce the risk of intergroup violence. Those tools have so far been used in a wide variety of countries, for two kinds of projects: studies of Dangerous Speech, i.e. collecting, classifying, and analyzing it; and efforts to diminish the harmful impact of such messages, for instance by teaching people to be less easily swayed by them.

Such work has been done in the Czech Republic, Ethiopia, India, Kenya, Myanmar, Nigeria, Pakistan, South Sudan, Sri Lanka, the United States - sometimes by the Dangerous Speech Project, and usually by other nonprofits or academics. For details of most extensive work, and relevant resources, see the links at left. Most of these countries are not at apparent risk of genocide or mass violence, but they are unfortunately also not free of Dangerous Speech.

Projects that collected and studied Dangerous Speech or hateful speech have been done in Ethiopia, Kenya, Myanmar, Nigeria, South Sudan, and Sri Lanka. Reports on all of them are easily available online. In Kenya the Umati project, conducted by iHub Research in 2012 and 2013, collected hateful and dangerous speech from online sources including Facebook, Twitter, blogs and newspaper comments. In Sri Lanka the Centre for Policy Alternatives produced Liking Violence: A Study of Hate Speech on Facebook in Sri Lanka in 2014. In Ethiopia, Mechachal: Online debates and elections in Ethiopia, was a collaboration between Addis Ababa University and the University of Oxford in 2015. In Myanmar, the Myanmar ICT for Development Organization (MIDO) monitored hate speech online, especially on Facebook, in 2015 and 2016. In Nigeria, The Centre for Information Technology and Development (CITAD), based in Kano, carried out a project to monitor hate speech and dangerous speech on social media in 2015 and 2016. Finally, in South Sudan, the Washington, D.C.-based PeaceTech Lab produced Social Media and Conflict in South Sudan: A Lexicon of Hate Speech Terms in 2016. As the title suggests, that project collected a set of what it identified as hate speech terms, not specific examples of their use.