Violence In Action Dataset (VIAD)

The Violence in Action Dataset (VIAD) is a unique dataset comprised of qualitative user-generated video material depicting real-world instances of war crimes including torture, the targeting of civilians, massacres, the mutilation of bodies, lynchings, and related material. The database is compiled from open source platforms of video material generated by perpetrators of these crimes, which have been distributed online for different reasons. VIPRE’s team has been constructing VIAD based on material dating back to 2011. To date, VIAD contains user-generated materials from war or conflict zones including Syria, Iraq, Lebanon, Egypt Ukraine, Russia, Colombia, Mexico, and Afghanistan, producing a dataset of over 1,000 videos. VIAD has been coded both qualitatively and quantitatively. Quantitatively, the material has been coded on a multitude of variables including the country of occurrence, the nature of the violence depicted, the identity of perpetrators, the emotional disposition of perpetrators, the number of perpetrators, the events preceding the crimes depicted, and the tools or technologies used. Qualitatively, the data has been coded through an ethnomethodological process tracing out the sequences of actions depicted at their minute corporeal, conversational, and other levels, in order to reveal specific material, emotional, affective, and aesthetic drivers of violence. Taken together, VIAD provides a uniquely granular set of data on the ‘just-whatness’ or ‘quiddity’ of violences that have – historically – rarely been seen by social scientists up close. The dataset was designed and created by Jonathan Luke Austin.

For ethical and legal reasons, VIAD is currently held under embargo until full anonymization procedures are applied and current legal debates over the possible uses of this material are clarified. Researchers interested in using VIAD in their own projects during this embargo period should contact Jonathan Luke Austin directly.