The research team built on the approach to the investigation of massacre set out by historical sociologist Jacques Semelin. He considers that the ‘triptych of aggressor, victim and witness, comprises the “basic triangle” for investigating massacre’. The investigator must examine the pre-conditions leading to the event and include oral and written accounts of the event at the time and in the aftermath (Semelin 2005: 376). He also notes that evidence produced in the long aftermath is often more reliable than in the immediate aftermath. From Semelin’s work, the research team established a definition of frontier massacre and its common features. Definition Unlike ‘genocide’, there is no legal definition of massacre, or a ‘frontier massacre’. Most international scholars of massacre appear to agree that the minimum number of people killed to constitute a massacre is between three and ten people (Dwyer and Ryan 2012: xiv-xv).