Chapter 3 expands on the experimental phase of the project, providing an account of its design and a thorough discussion of the results and their implications. The chapter explores the level of accuracy with which participants in IM are able to detect the substitution of one interlocutor with another, and the levels of confidence with which such decisions are made. It also addresses the effects of impersonator preparation on these scores. Finally, the linguistic criteria that people report having relied upon in making these assessments are scrutinised, and we are thus able to formulate opinions about which features are the most salient for the construction of one’s linguistic identity.