The Language of Control explores how people express threats and enforce agreements when formal institutions—such as courts or contracts—are weak or absent. In everyday settings like households, informal markets, or personal relationships, people regularly rely on verbal warnings and conditional statements (“If you do this, then I’ll do that”) to shape others’ behavior. Despite how common these interactions are, we know surprisingly little about how the phrasing of a threat—its tone, structure, or politeness—affects how it is interpreted and whether it succeeds in building trust, inducing compliance, or generating resentment.

The project combines behavioural economics and linguistics to study how informal threats are formulated and perceived across languages. During the fellowship, I will build a cross-linguistic dataset of informal threat statements in English, French, Spanish, Russian, and Hebrew. Participants are placed in a simple trust dilemma and asked to write or record what they would say to another person if that person behaved unfairly. The statement is then sent to a counterpart, who must decide how they would split the proceeds of the game after receiving the communication. The resulting messages are systematically analyzed for linguistic features such as conditional structure, modality, directness, and politeness.

In a second phase, participants evaluate anonymized threats and report how severe, credible, fair, or coercive they seem. Together, these experiments allow me to map specific linguistic features to how threats are interpreted across cultures. The project aims to produce a new framework for understanding informal enforcement—showing that threats are not simply present or absent, but vary in meaning and effectiveness depending on how they are expressed, the language in which they are delivered, and the cultural context in which they are deployed.