Deciphering the Emic Perspective in Data in Order to Assess Threat

ABSTRACT Methodologies and text analytic algorithms that consider the emic (or rstperson) perspective can help analysts interpret threat narratives, enabling them to better forecast violent events and prepare appropriate COAs. This chapter discusses the rationale, the approach, and some results for these methodologies and algorithms. It includes several real-world examples of their application, such as from the Islamic State and the Boko Haram, political and military actions between India and Pakistan, and competing actors in Philippines and Egypt. Empirical testing found that the emic discourse markers correlated with actual violence 69% of the time (much higher than the standard etic approaches); this increased to an 86% correlation when the emic and etic models were integrated.