Threatening is among the less pleasant "things we do with words", but, together with other conflictive speech acts, it seems to play a central role in communication. Yet, little is still known about how and when exactly speakers threaten. The present volume addresses this void by giving an in-depth analysis of the form and function of this speech act. A set of authentic threat utterances is used to probe questions on the linguistic repertoire employed and the different objectives speakers pursue with their threats. Based on the central findings, a classification of two types of threats is proposed, each with distinctive formal and functional properties. The analysis employs a mixed method approach with a two-fold aim; by combining a qualitative discussion of examples with the application of innovative statistical methodology, the findings allow new insights into research on threats and, simultaneously, offer new perspectives on general research methodology.