This study examines how speakers planning or generating reported speech and thought (RT), temporarily drop this production to incorporate other material; such inclusions recalibrate the wider RT context, and assure that the RT recipient is fit to read the speaker’s underlying interactional project.